CURVE BALLS

The essays I publish in ZAGROSCIENCE typically examine specific neuroscientific themes in tandem with corollary subject matters across non-science disciplines (e.g., philosophy, mythology, history, literature, etc.). For instance, in Harvest of Reality: The Cortical Organization of Belief, I explore the neural substrates of belief formation and revision while expounding on conditional probabilities and offering some commentary on general shifts in ideological positions throughout the covid-19 pandemic. In Echoes of a dying brain: Peering into the abyss through the lens of neuroscience, I discuss brain dynamics associated with near-death experiences, couching my review in an exposition on Leonid Andreyev’s thematically-salient novella, Lazarus. Cupid’s fingerprints in the brain: Exploring the neural basis of love delves into the cortical correlates of human bonding while additionally compiling a short list of poems written about love and affection spanning the last 4,000 years. In Cortical Wildfires, which relates to epilepsy and the research I conducted at the Montreal Neurological Institute, I consider Mesopotamian records, moon gods, and Euripedes’ tragedy of Herakles, which recounts the sad fate that awaited the mighty Greek demigod back home upon his triumphant return from his 12 labours. Once in a while, however, I like to throw my readership a curve ball just to spice things up! Durkistan: An Anagrammatic Odyssey Of Historical Genocide, which is a sci-fi tale with real-life earthly implications, is one of them. La passerelle cosmique, which is my first and only—but likely not last—surrealist French entry, is another.

I’ve had one particular essay sitting in my Substack draft list since August 9, titled, I’m Shahin, The Kurdish-Canadian man whom [REDACTED], which is the “curviest” of all my curve balls. With ~8.5K words, this piece is the longest I’ve written to date, and for good reason too: it is about how the [REDACTED] attempted to [REDACTED] me with [REDACTED]. It contains sensitive information about a soberingly life-changing experience I had back in February, which I have allusively described in previous articles as “kafkaesque.” The jury is still out on when or even whether I’ll go public with it, but I’m in the midst of serious talks with the powers that be to assess the suitability and appropriateness of its prospective publication. To be continued…

So, while we’re waiting, let me throw you a softer curve ball in this, the latest issue of ZAGROSCIENCE, where I’ll talk you through how I vibe-coded an entire MS-DOS style game called KAWA.EXE.

NOT MY FIRST RODEO

Believe it or not, back in 2020 during the global pandemic while I was still doing my PhD in cognitive neuroscience at McGill University, I actually developed a small combat game with myself as the playable character! Like many worthwhile things in life—including the inception of KAWA.EXE—the creation of this game was rather impromptu, and regardless of my hefty research workload, I still managed to make time for it because this pet project of mine simply appealed to me. Picking up random, transient callings, and following them to completion is one of my quirks, and I honestly think I’m intellectually and psychologically richer for it. Everything has a price, though, and the one I keep paying for has to do with obsessiveness, which can sometimes be off-putting to those around me…

Howbeit, after hours—if not days—of YouTube tutorials, Stack Overflow queries, and borderline pathological trial and error, I finally managed to build something that was not half bad—in the pre-vibe-coding era, that’s how coding worked! One thing my PhD studies taught me is that, I am—and in fact, we all are to some extent—much more resilient and perspicacious when failure is not an option. For instance, in during my first year, I went from having zero knowledge of the general-purpose programming language Python to fully designing, validating, and launching an entire Python-based functional MRI memory assessment software in less than 3 months. Here’s the GitHub link: https://github.com/MICA-MNI/micaopen/tree/master/task-fMRI

Below, you’ll find the full Python code—all 464 lines of it—I used in my combat game. For anyone who’s never seen programming syntax before, brace yourself, for what lies under the hood may be daunting. And for the expert programmers/developers/hackers among you (looking at you ), please forgive my bare-bones approach to commenting or any other coding faux pas you come across—I will one day atone for my many coding sins…

# housekeeping
import pygame

# initialize pygame
pygame.init()

# define clock
clock = pygame.time.Clock()

# define window
pygame.display.set_caption('whoop dat ass!')
pygame.mouse.set_visible(False)
win = pygame.display.set_mode((0, 0), pygame.RESIZABLE)
#win = pygame.display.set_mode((0, 0), pygame.FULLSCREEN)

# define background music & sound effects
bg_music = pygame.mixer.music.load('sound/bg_music.wav')
selectionSound = pygame.mixer.Sound('sound/selection.wav')
chooseSound = pygame.mixer.Sound('sound/choose.wav')
punchSound = pygame.mixer.Sound('sound/punch.wav')
whooshSound = pygame.mixer.Sound('sound/whoosh.wav')

# define screen displays
selection = [pygame.image.load('pix/char1.png'), pygame.image.load('pix/char2.png'), pygame.image.load('pix/char3.png')]
Pause = pygame.image.load('pix/pause.png')
Victory = pygame.image.load('pix/victory.png')
Loss = pygame.image.load('pix/loss.png')

# define functions for game mechanics
def pause():
    selectionSound.play()
    game_is_paused = True

    while game_is_paused:
        for ev in pygame.event.get():
            if ev.type == pygame.KEYDOWN:
                if ev.key == pygame.K_c:
                    game_is_paused = False
                elif ev.key == pygame.K_ESCAPE:
                    pygame.quit()
                    quit()
        pygame.mixer.music.pause()
        win.blit(Pause, (50, 250))
        pygame.display.update()

def victory():
    if not adnan.alive:
        pygame.mixer.music.stop()
        win.blit(Victory, (270, 150))
        pygame.display.update()
        pygame.time.delay(5000)
        pygame.quit()
        quit()

def loss():
    if not shoo.alive:
        pygame.mixer.music.stop()
        win.blit(Loss, (207, 150))
        pygame.display.update()
        pygame.time.delay(5000)
        pygame.quit()
        quit()

def redrawGameWindowW():
    temple.draw(win)
    adnan.draw(win)
    shoo.draw(win)
    pygame.display.update()

# define classes
class background():
    bg = [pygame.image.load('pix/background_1.png'), pygame.image.load('pix/background_2.png'),
          pygame.image.load('pix/background_3.png'), pygame.image.load('pix/background_4.png'),
          pygame.image.load('pix/background_5.png'), pygame.image.load('pix/background_6.png'),
          pygame.image.load('pix/background_7.png'), pygame.image.load('pix/background_8.png')]

    width, height = 1365, 780

    for i in range(8):
        bg[i] = pygame.transform.scale(bg[i], (width, height))

    def __init__(self):
        self.bgCount = 0

    def draw(self, win):
        self.bgCount += 1
        if self.bgCount + 1 >= 8:
            self.bgCount = 0
            win.blit(self.bg[self.bgCount], (0, 0))
        else:
            win.blit(self.bg[self.bgCount], (0, 0))

class hero():
    faceRight = pygame.image.load('pix/hero_1.png')
    walkRight = pygame.image.load('pix/hero_3.png')
    punchRight = [pygame.image.load('pix/hero_2.png'), pygame.image.load('pix/hero_2b.png'),
                  pygame.image.load('pix/hero_2.png'), pygame.image.load('pix/hero_2b.png')]
    jumpRight = pygame.image.load('pix/hero_4.png')
    takeHit = pygame.image.load('pix/hero_5.png')
    KO = pygame.image.load('pix/hero_6.png')
    heroStat = pygame.image.load('pix/shoo_stat.png')
    heroStat = pygame.transform.scale(heroStat, (175, 75))

    def __init__(self, x, y, l_extreme, r_extreme):
        self.x = x
        self.y = y
        self.path = [l_extreme, r_extreme]
        self.punchCount = 0
        self.velocity = 30
        self.jumpCount = 10
        self.hitCount = 0
        self.constant = []
        self.hitBox = (self.x + 50, self.y + 25, 92, 270)
        self.alive = True
        self.faceR = False
        self.faceL = False
        self.walkR = False
        self.walkL = False
        self.punch = False
        self.jump = False
        self.takeBlow = False
        self.deathFlag_R = False
        self.deathFlag_L = False
        self.postDead = []

    def hero_action(self):
        if not self.takeBlow:
            if not self.jump:
                if self.walkR:
                    if self.x + self.velocity < self.path[1]:
                        self.x += self.velocity
                        win.blit(self.walkRight, (self.x, self.y))
                    else:
                        win.blit(self.faceRight, (self.x, self.y))
                elif self.walkL:
                    if self.x - self.velocity > self.path[0]:
                        self.x -= self.velocity
                        win.blit(pygame.transform.flip(self.walkRight, True, False), (self.x - 57, self.y))
                    else:
                        win.blit(pygame.transform.flip(self.faceRight, True, False), (self.x - 57, self.y))
                elif self.faceR and not self.punch:
                    win.blit(self.faceRight, (self.x, self.y))
                elif self.faceR and self.punch:
                    if self.punchCount + 1 >= 4:
                        self.punchCount = 0
                    self.punchCount += 1
                    if adnan.hitBox[0] < self.hitBox[0] + self.hitBox[2] + 50 < adnan.hitBox[0] + 75:
                        adnan.take_Hit()
                        if adnan.hitCount > 168:
                            whooshSound.play()
                        else:
                            punchSound.play()
                            adnan.takeBlow = True
                    else:
                        whooshSound.play()
                    win.blit(self.punchRight[self.punchCount // 3], (self.x, self.y))
                elif self.faceL and not self.punch:
                    win.blit(pygame.transform.flip(self.faceRight, True, False), (self.x - 57, self.y))
                elif self.faceL and self.punch:
                    if self.punchCount + 1 >= 4:
                        self.punchCount = 0
                    self.punchCount += 1
                    if adnan.hitBox[0] + adnan.hitBox[2] - 75 < shoo.hitBox[0] - 50 < adnan.hitBox[0] + adnan.hitBox[2]:
                        adnan.take_Hit()
                        if adnan.hitCount > 168:
                            whooshSound.play()
                        else:
                            punchSound.play()
                            adnan.takeBlow = True
                    else:
                        whooshSound.play()
                    win.blit(pygame.transform.flip(self.punchRight[self.punchCount // 3], True, False), (self.x - 57, self.y))
            elif self.jump:
                if self.faceR:
                    if self.jumpCount >= - 10:
                        self.constant = 1
                        if self.jumpCount < 0:
                            self.constant = -1
                        self.y -= (self.jumpCount ** 2) * .8 * self.constant
                        win.blit(self.jumpRight, (self.x, self.y))
                        self.jumpCount -= 1
                    else:
                        win.blit(self.faceRight, (self.x, self.y))
                        self.jumpCount = 10
                        self.jump = False
                elif self.faceL:
                    if self.jumpCount >= - 10:
                        self.constant = 1
                        if self.jumpCount < 0:
                            self.constant = -1
                        self.y -= (self.jumpCount ** 2) * .8 * self.constant
                        win.blit(pygame.transform.flip(self.jumpRight, True, False), (self.x - 57, self.y))
                        self.jumpCount -= 1
                    else:
                        win.blit(pygame.transform.flip(self.faceRight, True, False), (self.x - 57, self.y))
                        self.jumpCount = 10
                        self.jump = False
        self.takeBlow = False

    def take_Hit(self):
        if adnan.hitCount < 168 and self.hitCount < 168:
            if not self.jump:
                if adnan.hitBox[0] < self.hitBox[0] + self.hitBox[2] < adnan.hitBox[0] + adnan.hitBox[2] or \
                        adnan.hitBox[0] < self.hitBox[0] < adnan.hitBox[0] + adnan.hitBox[2]:
                    self.takeBlow = True
                    if shoo.faceR:
                        self.hitCount += 12
                        win.blit(self.takeHit, (self.x, self.y))
                        punchSound.play()
                    elif shoo.faceL:
                        self.hitCount += 12
                        win.blit(pygame.transform.flip(self.takeHit, True, False), (self.x - 57, self.y))
                        punchSound.play()

    def get_KOed(self):
        if self.faceR and self.hitCount == 168:
            self.deathFlag_R = True
            self.takeBlow = True
        elif shoo.faceL and self.hitCount == 168:
            self.deathFlag_L = True
            self.takeBlow = True

        if self.deathFlag_R and not self.deathFlag_L and self.alive:
            if adnan.velocity < 0:
                win.blit(self.KO, (self.x - 75, self.y + 110))
                self.postDead = 1
            elif adnan.velocity >= 0:
                win.blit(pygame.transform.flip (self.KO, True, False), (self.x - 55, self.y + 110))
                self.postDead = 0
            self.alive = False
        elif not self.deathFlag_R and self.deathFlag_L and self.alive:
            if adnan.velocity < 0:
                win.blit(self.KO, (self.x - 75, self.y + 110))
                self.postDead = 1
            elif adnan.velocity >= 0:
                win.blit(pygame.transform.flip(self.KO, True, False), (self.x - 55, self.y + 110))
                self.postDead = 0
            self.alive = False

        if self.postDead == 1:
            win.blit(self.KO, (self.x - 75, self.y + 110))
        elif self.postDead == 0:
            win.blit(pygame.transform.flip (self.KO, True, False), (self.x - 55, self.y + 110))

    def draw(self, win):
        self.hero_action()
        self.take_Hit()
        self.get_KOed()
        adnan.get_KOed()
        win.blit(self.heroStat, (0, 0))
        if self.hitCount <= 168:
            pygame.draw.rect(win, (255, 0, 0), (4, 42, 168 - self.hitCount, 29), 0)
        self.hitBox = (self.x + 50, self.y + 25, 92, 270)
        # pygame.draw.rect(win, (255, 0, 0), self.hitBox, 2)

class enemy():
    walkRight = [pygame.image.load('pix/baddy1_1.png'), pygame.image.load('pix/baddy1_2.png')]
    takeHit = pygame.image.load('pix/baddy1_3.png')
    KO = pygame.image.load('pix/baddy1_4.png')
    enemyStat = pygame.image.load('pix/adnan_stat.png')
    enemyStat = pygame.transform.scale(enemyStat, (175, 75))

    def __init__(self, x, y, end):
        self.x = x
        self.y = y
        self.end = end
        self.path = [x - 1040, end]
        self.walkCount = 0
        self.velocity = -10
        self.hitCount = 0
        self.hitBox = (self.x + 25, self.y + 25, 75, 175)
        self.alive = True
        self.takeBlow = False
        self.deathFlag_R = False
        self.deathFlag_L = False

    def enemy_action(self):
        if self.hitCount < 168:
            if self.velocity > 0:
                if self.x + self.velocity < self.path[1]:
                    self.x += self.velocity
                else:
                    self.velocity = self.velocity * -1
                    self.walkcount = 0
            else:
                if self.x - self.velocity > self.path[0]:
                    self.x += self.velocity
                else:
                    self.velocity = self.velocity * -1
                    self.walkcount = 0

    def take_Hit(self):
        if self.hitCount < 168:
            if shoo.faceR and shoo.punch:
                self.hitCount += 24
                win.blit(pygame.transform.flip(self.takeHit, True, False), (self.x, self.y))
            elif shoo.faceL and shoo.punch:
                self.hitCount += 24
                win.blit(self.takeHit, (self.x, self.y))

    def get_KOed(self):
        if shoo.faceR and self.hitCount == 168:
            self.deathFlag_R = True
            self.hitCount = self.hitCount + 1
        elif shoo.faceL and self.hitCount == 168:
            self.deathFlag_L = True
            self.hitCount = self.hitCount + 1

        if self.deathFlag_R and not self.deathFlag_L:
            win.blit(pygame.transform.flip(self.KO, True, False), (self.x, self.y + 110))
            self.alive = False
        elif not self.deathFlag_R and self.deathFlag_L:
            win.blit(self.KO, (self.x, self.y + 110))
            self.alive = False

    def draw(self, win):
        self.enemy_action()
        # self.get_KOed()
        if self.hitCount < 168 and not self.takeBlow:
            if self.walkCount + 1 >= 6:
                self.walkCount = 0

            if self.velocity > 0:
                win.blit(self.walkRight[self.walkCount // 3], (self.x, self.y))
                self.walkCount += 1
            else:
                win.blit(pygame.transform.flip(self.walkRight[self.walkCount // 3], True, False), (self.x, self.y))
                self.walkCount += 1
        self.takeBlow = False
        win.blit(self.enemyStat, (1190, 0))
        if self.hitCount <= 168:
            pygame.draw.rect(win, (255, 0, 0), (1194 + self.hitCount, 42, 168 - self.hitCount, 29), 0)
        self.hitBox = (self.x + 25, self.y + 25, 75, 175)
        # pygame.draw.rect(win, (255, 0, 0), self.hitBox, 2)

# define class instances
shoo = hero(100, 460, 15, 1140)
adnan = enemy(1140, 540, 1140)
temple = background()

# run game
game_is_running = False
char_sel = True
arrow_count = 0
char1 = False
char2 = False
char3 = False
whoop_dat_ass = True

while char_sel:

    for event in pygame.event.get():
        if event.type == pygame.KEYDOWN and event.key == pygame.K_ESCAPE or event.type == pygame.QUIT:
            pygame.quit()
            quit()
        elif event.type == pygame.KEYDOWN and event.key == pygame.K_RETURN:
            char_sel = False
            game_is_running = True
            if arrow_count == 0:
                char1 = True
            elif arrow_count == 1:
                char2 = True
            elif arrow_count == 2:
                char3 = True
            pygame.display.update()
            chooseSound.play()

        if arrow_count == 0:
            if event.type == pygame.KEYDOWN and event.key == pygame.K_DOWN:
                arrow_count += 1
                selectionSound.play()
        elif arrow_count == 1:
            if event.type == pygame.KEYDOWN and event.key == pygame.K_DOWN:
                arrow_count += 1
                selectionSound.play()
            elif event.type == pygame.KEYDOWN and event.key == pygame.K_UP:
                arrow_count -= 1
                selectionSound.play()
        elif arrow_count == 2:
            if event.type == pygame.KEYDOWN and event.key == pygame.K_UP:
                arrow_count -= 1
                selectionSound.play()

        if arrow_count == 0:
            win.blit(selection[0], (0, 0))
        elif arrow_count == 1:
            win.blit(selection[1], (0, 0))
        elif arrow_count == 2:
            win.blit(selection[2], (0, 0))

        if char1:
            pygame.draw.rect(win, (178, 34, 34), (430, 150, 505, 200), 10)
        elif char2:
            pygame.draw.rect(win, (178, 34, 34), (430, 352, 505, 200), 10)
        elif char3:
            pygame.draw.rect(win, (178, 34, 34), (430, 557, 505, 200), 10)

        pygame.display.update()

pygame.time.delay(1000)

while whoop_dat_ass:

    for event in pygame.event.get():
        if event.type == pygame.KEYDOWN and event.key == pygame.K_ESCAPE or event.type == pygame.QUIT:
            pygame.quit()
            quit()
    whoopAss = pygame.image.load('pix/whoop_dat_ass.png')
    win.blit(whoopAss, (0, 0))
    pygame.display.update()
    pygame.time.delay(2000)
    whoop_dat_ass = False

pygame.mixer.music.play(-1)
shoo.faceR = True
shoo.faceL = False

while game_is_running:

    clock.tick(18)

    shoo.punch = False
    shoo.walkR = False
    shoo.walkL = False

    for event in pygame.event.get():
        if event.type == pygame.KEYDOWN and event.key == pygame.K_ESCAPE or event.type == pygame.QUIT:
            pygame.quit()
            quit()
        elif event.type == pygame.KEYDOWN and event.key == pygame.K_p:
            pause()
            pygame.mixer.music.unpause()

    key = pygame.key.get_pressed()

    if shoo.alive:

        if key[pygame.K_RIGHT]:
            shoo.walkR = True
            shoo.walkL = False
            shoo.faceR = True
            shoo.faceL = False
            shoo.punch = False

        elif key[pygame.K_LEFT]:
            shoo.walkR = False
            shoo.walkL = True
            shoo.faceR = False
            shoo.faceL = True
            shoo.punch = False

        elif key[pygame.K_q]:
            shoo.punch = True

        elif key[pygame.K_UP]:
            shoo.jump = True

    if not shoo.alive or not adnan.alive:
        game_is_running = False

    redrawGameWindowW()

victory()
loss()

BLACKTHORNE—MY FIRST GAMING LOVE AFFAIR

As you may have gathered by now, I’m a geek, and any geek worth his salt loves video games! I admit I haven’t played one for over a year—it’s a bit tough when you’re a husband and a dad with grown-up responsibilities—but I’ve had my share of make-believe adventures. Here’s a list of video games I’ve played throughout the years that span different genres, which I’d recommend to anyone (items marked in bold are my favourites):

• Commandos: Behind Enemy Lines (1998)

• Commandos 2: Men of Courage (2001)

• Commandos 3: Destination Berlin (2003)

• Batman: Arkham Asylum (2008)

• Limbo (2010)

• Batman: Arkham City (2011)

• Batman: Arkham Origins (2013)

• Batman: Arkham Knight (2015)

• Metal Gear Solid V: The Phantom Pain (2015)

• The Witcher 3: Wild Hunt (2015)

• Inside (2016)

• God of War (2018)

• Red Dead Redemption 2 (2018)

• Ghost of Tsushima (2020)

• Elden Ring (2022)

• God of War: Ragnarök (2022)

I’ve really been meaning to play the latest 2026 hit Crimson Desert, but I actually have a life and don’t want to lose touch with reality—on account of the previously mentioned grown-up responsibilities—which is what will likely end up happening if I go anywhere near it! There is one game, however, which will forever be my original love, not because it’s necessarily the best game I ever played—in fact, I never even made it past the first few levels—but simply by virtue of being my first MS-DOS gaming experience as a child, Blackthorne (1994).

Blackthorne is a sci-fi/fantasy/2D shooter in which the protagonist, Kyle—yes, that’s his name—must return to his realm in Androth to reclaim his rightful throne from the usurper Sarlacc. It’s got shotguns, glowing bombs, metal gates, ogre-like henchmen, innocent captives, boss fights, and—something I particularly enjoyed as a kid for some reason—levitating platforms that make a neat whooshing sound! One day, I will go back to this game and play it to completion. But for now, let me tell you about KAWA.EXE, a MS-DOS style game I whimsically started developing using AI.

SPUR-OF-THE-MOMENT

Since January, I’ve been tirelessly working on my start-up, We Are Philanthropy Inc., which I’m hoping to launch before 2027—and I’ll have much more on this exciting new initiative in the months to come! Waiting for the related legal documentation to be approved, I found my wondering mind randomly stumbling upon one thought: “Wouldn’t it be cool if I could make a game like Blackthorne?” And then it hit me: “I probably could with AI assistance!” And just like that, with one simple prompt in Claude, KAWA.EXE was born.

I honestly wasn’t expecting Claude to actually create a functioning video game right off of the bat, but it did, and I was hooked. Unsurprisingly, the first iteration was rather crude, requiring several rounds of optimization. But before sharing with you the method behind my madness, let’s first turn our attention to the protagonist of the game, Kawa, and his story.

KAWA—THE BLACKSMITH

In Kurdish mythology, Kawa is the heroic blacksmith who delivered his people from the evil ruler Zahhak. As the legend goes, the tyrant was afflicted with an unusual condition in which two serpents grew from his shoulders, causing Zahhak much agony. To assuage his insufferable pain, the malevolent king required daily human sacrifices to satiate his ravenous glenohumeral serpents. Kawa lost his children in this wretched manner, and so he vowed to put an end to Zahhak’s unjust rule once and for all. One day, Kawa scaled the mountains to reach Zahhak’s throne and defeated him in battle, terminating his despotic reign. To usher in the new era, the heroic blacksmith lit the fire of Newroz/Nuroj, which is Kurdish for “new day.”

The 9^th century Kurdish polymath, Abu-Hanifa Dinawari, attributed the origin of the Kurds to this story in his Al-Akhbār al-ṭiwāl (“The Long Narratives”):

They say: every day he brought four stout men, and they were slaughtered and their brains were taken, and the two snakes were fed with them. He [Zahhak] had a vizier from among his folk, and he put in charge of his ministry a man from the sons of Arfaḫšađ, who was called Armiyāyīl. When he brought the men to be slaughtered, he spared the life of two of them, and put in their stead two rams. And he ordered the two men to go where there would be no trace of them. They went into the mountains, and they stayed in them and did not come near villages or cities. It is said that they were the origin of the Kurds.

Various forms of this story have been told throughout history, with the Shahnameh being arguably its most well-known version, though it was compiled and finalized in the early 11^th century—over a century after Dinawari’s account—by Abu al-Qasim Mansur bin Hasan Tos, better known by his pen name, Ferdowsi. Interestingly, Dinawari’s narrative makes no mention of Kawa, attributing his exploits to Nimrod instead:

He [Nimrod] came to them [sons of Arfaḫšađ] and they appointed him king over them. And he lorded it over those in the land of Bābil among the folk of the house of Ḍaḥḥāk [Zahhak], and then he killed them all together and took possession of the kingdom of Ḍaḥḥāk. [News of] this reached Ḍaḥḥāk and he headed toward him, but Numrūđ [Nimrod] defeated him and struck him upon the nape of the neck with a sharp mace. He felled him and bound him with a shackle, and took him to a cave in the mountain of Dunbāwand. He put him in it and sealed it. Numrūđ’s rule was strengthened and consolidated.

According to historian Soran Hamarash, the narrative pillars underlying the myth of Nimrod/Kawa and Zahhak may be rooted in the ancient Sumerian story of Ninurta and Asag, in which the former, a heroic Mesopotamian god, defeats the latter, a hideous demon of the mountains, with his magical mace called Sharur. Hamarash identifies Nimrod/Kawa and Zahhak as contemporaneous archetypes of Ninurta and Asag, respectively. My own view is that Zahhak bears resemblance to another Sumerian deity, Ningishzida; the fact that they both have snakes on their shoulders is a little too on the nose.

Today, these myths and stories constitute core elements in the national identity of Greater Kurdistan. In the Yarsanî/Kakeyî creed, for instance, these tales live on through oral and written traditions known as “kelam,” oftentimes recited in ritualistic performances during which singer-musicians chant in chorus while playing their sacred Kurdish tembûr—a fretted, three-stringed musical instrument. See the video below:

Here’s a lyrical excerpt from the video [15:00-15:41]:

Dwanze ferzenan Kawey ahenger— (x2)

[The twelve children of Kawa the blacksmith]

Yekî çeneşmen, Kawem, Kawem—

[Only one is left, my Kawa, my Kawa]

Yanzeş birrîn ser, gyan ey gyan ey gyan.

[And eleven were decapitated, dear oh dear oh dear]

METHOD BEHIND MADNESS

The first iteration of KAWA.EXE was quite rudimentary, with blob-like sprites and bland, polygonal backdrops. So, my first thought was to build models for Kawa and Zahhak, as well as a generic one for Zahhak’s foot soldiers. While I use Claude for the actual HTML game code, I find Gemini to be far superior for visual design. After a few rough drafts, I settled on the following sprites and their associated states:

I repeated the same basic iterative prompting procedure for level designing, resulting in the following arenas (yes, they’re somewhat derivative, but it’s ok…):

I then created a few narrative exposition slides to provide some contextual background and set the tone of the game:

Finally, I asked Claude to compose a very basic background music inspired from Middle Eastern scales, and that was pretty much it! No YouTube tutorials, no Stack Overflow queries, no borderline pathological trial and error on my behalf… I’m not saying it’s better one way or the other—I’m simply making an observation on just how much AI has changed the playing field.

PLAY KAWA.EXE NOW

Click the link below to start playing KAWA.EXE now! Don’t be vexed by the odd looking URL—it’s an automatically-generated web address by Netlify, the deployment platform that is hosting the HTML file for the game. Bear in mind that while you can access the website on your smart phone, you’ll need to use a keyboard to actually play (e.g., ←→ to move left/right; ↑ / Z to jump; SPACE to hammer attack; ENTER to start/continue; M to mute/unmute music).

Don’t expect KAWA.EXE to win game of the year anytime soon; it is extremely unsophisticated even by MS-DOS standards. But, it is still quite extraordinary how AI enabled me to build a fully functional video game in an hour or so using simple prompts and the power of my own imagination!

CHALLENGE

So, I now challenge you, dear reader, to create your own game using the basic steps I’ve shared with you in this piece—show me your “mettle” and see if it can hold a candle to my forging hammer!

Subscribe now

REFERENCES

1. Bonner, M. R. J. (2014). An historiographical study of Abū Ḥanīfa Aḥmad ibn Dāwūd ibn Wanand al-Dīnawarī’s Kitāb al-Aḫbār al-Ṭiwāl (especially of that part dealing with the Sasanian kings) [Doctoral dissertation, University of Oxford].

2. Hamarash, S., 2022. The Lost and Untold History of the Kurds: Rediscovering the Beginning of the Western Civilisation and the Origin of the Indo-European Languages, ca. 10000 BC- 1300 CE. Iraq: Self-published.

3. Izady, M. R. (2018, October 9). Abu-Hanifa Ahmad Dinawari. Kurdistanica. https://kurdistanica.com/282/abu-hanifa-ahmad-dinawari/

4. Ningishzida. (2025, April 11). In Wikipedia. Retrieved August 22, 2026, from https://en.wikipedia.org/wiki/Ningishzida

In my piece dating back to March (see https://zagroscience.substack.com/p/substack-offered-me-a-faustian-bargain), I offered a mea culpa regarding my absence from Substack, which read:

Early that month [February], through no fault of my own, I was hurled into a living nightmare, a Kafkaesque reality that is still looming about, which effectively upended some of my priorities in life. When the time is right, I will go public with my story to raise awareness about what my family and I have had to endure in the hopes that no one else suffers the same fate […]

Although I still hesitate to disclose the nature of the events in question, I can say that the worst is finally behind us and, for the first time since early February, my family and I are breathing a sigh of relief. I should also note that I have an entire article—which reads more like a novella—written about these events, and ready for publication at a moment’s notice. I’m biding my time…

In that same piece, I also noted the following:

As it turns out, throughout these difficult times, most of my cognitive bandwidth has been taken up by a grand project that I started about two months ago [January], which, I’m hoping, will lay the groundwork for how social media is used in the future—and I’ll be elaborating more on this initiative as the year progresses.

This project is still ongoing and has reached new milestones. I’m really hoping to be able to divulge everything before the year’s end, but I’ll play it close to the chest for the time being.

WHY AM I EVEN TALKING ABOUT MARTIAL ARTS?

For starters, this wouldn’t be my first time talking about martial arts on ZAGROSCIENCE (see https://zagroscience.substack.com/p/unlocking-new-abilities-lessons-from). As a lifelong practitioner, if I had to choose which label best describes me between (a) neuroscientist and (b) martial artist, I’d honestly go with (b). A more nuanced answer would be: a martial artist who became a neuroscientist…and then an entrepreneur! But beyond defining a huge aspect of who I am, I’m talking about martial arts in this piece because, thanks to my neuroscience pedigree, I possess the tools to use myself as a lab rat in what promises to be an interesting experiment with N = 1. Indeed, I’ve devised a training program that I’ll be sticking to for 12 weeks in which I’ll be attempting to improve my fighting skills without ever doing a single round of sparring against another opponent—and in this issue of ZAGROSCIENCE, I’ll let you in on my secret weapon: mental imagery…

KASHIWAZAKI: THE LONELY RONIN

A few years ago while I was still living in Montreal, my Sifu mentioned a Japanese Judoka who’d train alone, show up to tournaments, and mop the floor with everyone! The individual in question was Katsuhiko Kashiwazaki, a multiple times judo champion, including the 1981 World Judo Championships gold medalist in the men’s 65kg category. For anyone who knows anything about judo and combat sports in general, you’ll know that to hone your fighting skills in any meaningful way, you need to train regularly against another person in a live grappling match or sparring session. So, I was genuinely taken aback by Sifu’s story, that a judoka who trained solo could, not only improve his craft, but effectively achieve some of the highest accomplishments in the sport. I catalogued the story in mind as an intriguing oddity, where it remained dormant until recently.

Then, in 2023, my family and I moved to Ottawa, where I worked on completing the final year of my PhD studies remotely, drafting my dissertation in preparation for my thesis defence in December of the same year at the Montreal Neurological Institute. Throughout this time, my wife and I were also caring for our baby daughter, so my circumstances weren’t ideal as far as martial arts training goes (NB: changing diapers while rehearsing for your PhD oral defence is quite the experience!). Between then and now, a lot of inertia set in, and while I’ve maintained a good weight training routine, I’ve only recently resumed training martial arts again, doing mostly katas, combat drills, and sparring combinations, all alone. So, it’s safe to say I haven’t done any actual combat against another opponent for over 3 years!

Earlier this month, I went back to Montreal and met up with Sifu again. I don’t know if it was the coffee he gave me, what he was telling me, just being in his presence, or a combination of all those things, but a fire was lit under me! I needed to get back into martial arts training as soon as possible. The Japanese Judoka ipponed back into my mind! As it turns out, one of my old juniors, Emery, who’s a radiologist now and an accomplished martial artist (e.g., black belt in Kudoki and Judo), had actually attended a seminar in Montreal taught by Kashiwazaki, himself (see video below sent to me by Emery, in which Kashiwazaki demonstrates tomoe nage, colloquially known as the James Bond throw). I learned from him that Kashiwazaki was already an accomplished judoka prior to training solo. Apparently, after leaving Tokyo, he had settled in a rural area where he didn’t have access to training partners, so he was forced to practice alone. I told Emery that, without knowing anything else about Kashiwazaki, I’d bet he incorporated some form of mental training into his solo practice. Indeed, I wasn’t wrong. It turns out that technique visualization was an integral part of his individual training regimen.

Reflecting a little deeper on the matter, I realized that non-contact training for actual professional combat sports athletes is not so uncanny. Indeed, there are high profile mixed martial arts (MMA) fighters who have openly stated that, at various points in their careers, they had either completely forgone sparring or reduced it significantly while still competing at the very highest level. The likes of “Ruthless” Robbie Lawler, “Mighty Mouse” Demetrious Johnson, and Max “Blessed” Holloway come to mind. Perhaps, then, it is possible to hone your fighting prowess without actually having to fight all the time…

TESLA: THE LATENT INVENTOR

I found inspiration in the least likely place, My Inventions, a compilation of articles written by the great Nikola Tesla. I was immediately absorbed by his self-professed capacities for mental imagery. Growing up, he had very few friends, so to remedy his loneliness, Tesla would spend hours lost in his imagination, which he described as being a living construct that is stupendously detailed and vivid. Indeed, his aptitude to see through the mind’s eye was truly remarkable—and arguably his greatest asset in devising his inventions. In one excerpt, he remarked:

When I get an idea I start at once building it up in my imagination. I change the construction, make improvements and operate the device in my mind. It is absolutely immaterial to me whether I run my turbine in thought or test it in my shop. I even note if it is out of balance. There is no difference whatever, the results are the same. In this way I am able to rapidly develop and perfect a conception without touching anything. When I have gone so far as to embody in the invention every possible improvement I can think of and see no fault anywhere, I put into concrete form this final product of my brain. Invariably my device works as I conceived that it should, and the experiment comes out exactly as I planned it. In 20 years there has not been a single exception.

The Japanese Judoka again nage-waza-ed back into my mind! He, too, like Tesla, had a predilection for mental imagery, but instead of implementing it to tinker with his cognitive contraptions, Kashiwazaki used it to perfect his art of throwing people on their donkeys! What if I, myself, as a black belt with decades of martial arts training and a fairly well consolidated mental library of tried and tested combat techniques used the same training modality to brush up on some old skills? That’s when it hit me: can I actually train alone and get better as a fighter by incorporating in my routine mental imagery training?

WHAT IS MENTAL IMAGERY ANYWAY?

Mental imagery, also known as motor imagery, is when you run a mental simulation of an action without actually executing it physically. It is a complementary modality to physical training, oftentimes used to reinforce learned mechanical patterns and grant athletes mental readiness during competition. Sports psychologists implement it in their practice via a framework known as PETTLEP, which is an acronym for Physical, Environmental, Task, Timing, Learning, Emotion, and Perspective. Here’s how mental imagery is performed through this approach:

• Physical: recreate the physical dimension of the activity/competition (e.g., tennis player holding a racket, boxer wearing trunks, bodybuilder getting a spray tan, etc.).

• Environmental: imagine the exact location of the activity/competition (e.g., tennis player visualizing the tennis court, boxer imagining the ring, bodybuilding picturing the stage, etc.).

• Task: focus on a specific mechanical pattern (e.g., the serve, the jab, the “most muscular” pose, etc.).

• Timing: synchronize the mental simulation of the task with how long it actually takes to do it physically (e.g., don’t run the 100-meter-dash in 5.25 seconds in your mind unless you can actually do it!).

• Learning: adapt and evolve in light of new information, skill acquisition, correction, etc.

• Emotion: incorporate the psychological state of mind associated with the activity/competition (e.g., exhilaration, anxiety, fear, determination, etc.).

• Perspective: select the optimal frame of reference for the task (e.g., first-person “egocentric” vs. bird’s eye view “allocentric”).

NB: while I don’t specifically mention it during my discussion of the scientific literature, many of the studies I cover in this piece incorporate PETTLEP.

WHAT DOES THE SCIENTIFIC LITERATURE SAY?

From a neuroscientific perspective, visualizing something (e.g., seeing it in your mind’s eye) and actually observing the thing itself in realtime share a lot of the same neural signatures. In a 2015 scientific review titled Mental Imagery: Functional Mechanisms and Clinical Applications, Pearson et al. highlighted undeniable parallels between the functional substrates of mental imagery and actual perception, suggesting that the former is a weaker and noisier version of the latter (see Fig. 1). Indeed, both modalities exhibit overlapping cortical correlates from the lowest levels of visual processing (e.g., V1: primary visual cortex) all the way up to higher areas (e.g., ventral temporal lobe). Findings suggest that SNR (signal-to-noise ratio) from mental imagery increases as information flows up the visual hierarchy. In fact, neural activation patterns have been used by multivariate pattern classifiers (MVPCs) to decode visual contents in both mental imagery and visual perception experimental paradigms, with more robust findings in higher level regions (see https://zagroscience.substack.com/p/multivariate-pattern-analysis for a discussion on MVPCs).

The authors of the paper stated:

Thus, MVPC studies of visual mental imagery support the claim that patterns of activity in perception and imagery become increasingly similar with ascension of the visual hierarchy […]

If anything, this review speaks to an implicit truth about how the human brain works: internally- and externally-driven subjective experiences differ in their origins, but at the end of the day, they’re both subjective experiences all the same. And this is where I’ll allow myself a small but relevant digression as regards the breakthrough psychedelic trip: anyone who’s had one will swear that what they’ve experienced felt every bit as real as normal reality—and in some cases even more real (see https://zagroscience.substack.com/p/chaos for a discussion on classic psychedelics)! Assuming that these substance-induced journeys aren’t some kind of physical boundary shattering event in which you are literally transported into a latent space of existence—I don’t even know what to believe anymore—then it stands to reason that some types of internally generated experiences can, in fact, prove profoundly immersive. I’m of the persuasion that psychedelics significantly boost the SNR of low level, spontaneous internal triggers, leading to astonishingly crisp mental images that take form as if they were being unveiled before your very own eyes. What if mental imagery training could help you enhance the strength and dampen the noise of self-generated visual constructs in a similar way? If my hunch is right, through training, you could potentially hone such cognitive skills so as to enable yourself to be immersed in your own internally-driven simulations. Imagine that, a tailor-made virtual reality system inside your head!

But wait, there’s more to the neuroscience than meets the eye! Mental imagery in which you are not simply imagining an object, but actually conducting a mechanical action is believed to recruit regions of the brain involved in movement execution. And while it’s true that it does, indeed, tap into the motor processing system, there’s nuance in how mental imagery and actual movement execution are associated with respect to neural representation. A widely accepted framework known as Motor Simulation Theory holds that imagining an action and physically executing it engage nearly identical cortical regions (e.g., prefrontal cortex [task goal], premotor cortex [action selection], SMA/pre-SMA [sequence planning], parietal cortex [body representation], basal ganglia [movement fine-tuning], cerebellum [coordination], primary motor cortex [execution]), the only difference between the two being that muscular activation, which is responsible for mechanical output, is neurologically inhibited in one, but not the other.

And still, the plot has thickened in recent years! A new meta-analysis by Van Caenegem et al. (2026) conducted a quantitative comparison between the neural representations of three cognitive states (e.g., “motor” imagery, movement execution, and working memory) across 697 neuroimaging studies. Using activation likelihood activation (ALE), a meta-analytic algorithm applied to brain activation data to determine the significance of convergence across different experiments, found that motor imagery (e.g., mental imagery) showed greater network overlap with working memory than it did with movement execution. While control analyses showed comparable activation profiles across mental states (e.g., motor imagery & working memory: 36% overlap; motor imagery & movement execution: 32% overlap), authors argued that emerging theories like the Motor-Cognitive Model offer greater explanatory power over previous frameworks, suggesting that mental imagery may occupy a space between motor and central executive systems (see Fig. 2).

Beyond neuroscience, I also came across a paper by Slimani et al. (2017) titled Effects of mental training on muscular force, hormonal and physiological changes in kickboxers, the theme of which is, as you can tell, very much aligned with my own objectives. Over the course of 12 weeks, researchers followed 53 young male kickboxers with a minimum of 4 years of training, randomly assigning them to 3 groups: (1) physical-training group (PG: N = 20), (2) physical & mental training group (PMG: N = 18), and (3) control group (CG: N = 10). PG partook in a regimen consisting of resistance training, plyometric drills, kickboxing sessions, and neutral cognitive tasks during rest. PMG, in addition to the same physical program as PG, also performed a mental training package (MTP), which comprised of a mix of motivational self-talk and internal mental imagery (e.g., imagining performing kickboxing techniques from a egocentric perspective). Neither physical nor mental training was done in CG. Unsurprisingly, both experimental cohorts saw great improvements with respect to the control after the 12-week period, but the real story were the group differences between PMG and PG, where PMG showed significant enhancements in muscular strength and upper body power, while displaying notable improvements in overall hormonal levels (e.g., testosterone, cortisol, & testosterone-to-cortisol ratio — see Fig. 3) and physiological markers (e.g., blood pressure & resting heart rate). The authors identified mental imagery training as an important component in the reduction of hormonal and physiological stress, which, in turn, were believed to drive observed muscular benefits.

A more recent systematic review by Andreato et al. (2022) offered additional insights into the effects of general mental training on a series of combat-relevant outcome measures across various martial arts. Titled What do we know about the effects of mental training applied to combat sports? A systematic review, the paper examined 21 studies with a total of 417 subjects, including 8 combat sports/martial arts (e.g., Karate, Judo, Taekwondo, Wrestling, Kickboxing, Brazilian Jiu-jitsu, MMA, & Wushu). Outcome measures fell into one of two primary categories: psychological measures (e.g., anxiety, self-confidence, mood/emotion, cognitive/social skills, & psychological well-being) and physical/athletic parameters (e.g., strength/power, reaction time, balance, & general performance). Across studies, mental training was implemented in a number of ways (e.g., mental imagery, self-talk, neurofeedback, mindfulness, etc.), which varied in duration (e.g., 4 minutes to 120 minutes), frequency (e.g., 2 sessions/week to daily), and delivery (e.g., isolation vs. combined/mixed “packages”).

The review indicated that mental training can help enhance combat-relevant skills by improving “anxiety, mood, self-confidence, emotions, cohesion, emotional balance, and relaxation, in addition to physical and athletic performance.” It also noted that implementing “packages” (e.g., mental imagery + motivational self-talk) may be more beneficial than conducting mental training modalities in isolation. While findings hold promise for the field of sports psychology in general and its applications to combat sports in particular, the authors also cited important research hurdles that must be overcome before drawing concrete conclusions, such as limited high quality studies that adequately account for randomization, control, and blinding, as well as a dearth of experiments conducted in female athletes.

An even newer multilevel meta-analysis by Liu et al. (2025) investigated the impact of mental imagery on athletic performance by systematically addressing variations in training parameters, such as session load, weekly frequency, and total program duration. Synthesizing data from 86 randomized controlled trials with over 3,500 athletes, the authors determined that imagery practice can significantly enhance performance outcomes, especially in agility and muscular strength. Researchers also noted that integrating other psychological skills training modalities (e.g., self-talk, video observation, virtual reality, mindfulness, etc.) into the mental imagery regimen further boosts its effectiveness. For best results, the paper recommended an optimal schedule consisting of 10-minute-long sessions performed thrice weekly over the course of 100 days (see Fig. 4). The study provided empirical evidence to help refine the psychological aspect of athletic programs while recognizing the need for further research given current limitations.

WHAT ABOUT CARDIO?

Cardio is the first casualty of inactivity, and I’ve learned this truth the hard way. There’s nothing more frustrating to a fighter than gassing out mid-sparring session, huffing and puffing, and feeling like you just inhaled a cup of sand. Many a great fighter has lost a bout because of poor cardio. On the flip side, many a mediocre fighter has won on the merits of a seemingly inexhaustible gas tank. Put technical proficiency and unrelenting cardiovascular fitness together, and what do you get it? Well, you get the apex fighter, the likes of Georges St-Pierre (GSP), who’s also a fellow Montrealer! So, I made a pact with myself: if anyone is going to beat me in sparring, it’d have to be because they’re technically and reactively better than I am—not because my cardiac output couldn’t match my physical intensity. So, I conducted another scientific literature review, this time, on how best to boost my VO₂max, which is a measure of how efficient your body is at consuming oxygen during physical exertion.

The 2024 systematic review/meta-analysis done by Thurlow et al. investigated a single exercise modality I’d heard about before: repeated-sprint training (RST). Its twofold objectives were: (1) examine the effectiveness of RST on a number of performance-related outcome measures (e.g., VO₂max, recovery, countermovement jump, etc.); (2) assess the modulatory effect of RST program parameters (e.g., duration, frequency, volume, etc.) on these markers of athletic performance. A total for 40 scientific publications were included, with 48 RST groups (N = 541) and 19 active control groups (N = 213). Authors reported an average 4.0% improvement above baseline in aerobic capacity for the RST groups (e.g., +2.2mL/(kg·min)). This effect size means that, after the training regimen, RST athletes could consume an extra ~2.2mL of oxygen per kilogram of body weight over the course of one minute (see Fig. 5). Interestingly, active controls actually regressed (e.g., -1.2mL/(kg·min)).

As regards program variables, the study defined an optimal regimen consisting of 3 cycles of 6 sets of 30 meters of straight-line sprints with 20-30 seconds of rest between sets and 4 minutes of rest between cycles, performed twice weekly for 6 weeks. Modifying these parameters beyond this reference was found to either reduce athletic performance or yield negligible benefits. So, compared to active control cohorts, RST groups did significantly better in cardiovascular output, especially if their regimens didn’t deviate much from the reference program. But how does RST compare to other tried and tested protocols designed specifically to improve VO₂max?

Yang et al. (2025) offered some important insights into this question. Titled Comparison of different interval training methods on athletes’ oxygen uptake: a systematic review with pairwise and network meta-analysis, the paper presented interesting findings pertaining to a three-way comparison between high-intensity interval training (HIIT), sprint interval training (SIT), and RST, with continuous training (CT) and conventional training (CON) serving as double controls. Analyzing data from 51 studies with a total of 1,261 athletes, a probabilistic ranking was computed with the following hierarchy: RST > HIIT > SIT > CT ≥ CON (see Fig. 6). While the main modalities (e.g., HIIT, SIT, and RST) showed considerable improvements in VO₂max over controls, differences between them were insignificant. Even so, what the probabilistic ranking (e.g., P-score) entails is that, if you were to bet on which modality is the likeliest to offer the absolute highest yield in VO₂max, RST would be your best wager. I had all the scientific rationale I needed now: RST it is!

MY GRAND SYNTHESIS

With the science-based insights absorbed, it was now time to design the two core modules (e.g., mental imagery & cardiovascular) of my program. Fortunately, hebdomadal best practices for both components dovetailed harmoniously, with 3 sessions per week optimized for mental imagery training, and 2 for RST. My weekly partition was a no-brainer: mental imagery module on Monday, Wednesday, and Friday (or Saturday), and cardiovascular module on Tuesday and Thursday.

With the preliminary outline of my program sketched out, I contacted Sifu and asked him for vital technical advice on the inclusion of combat fundamentals into my burgeoning protocol. I define basics/fundamentals as those techniques that confer to the combatant the highest rates of success. With Sifu’s help, I came up with the following:

• Striking techniques: straight punch (e.g., jab), hook punch, uppercut, front kick, roundhouse, and knee strikes.

• Grappling techniques: o-goshi, harai-goshi, uki-goshi, de ashi barai, ko-uchi gari, and kote-geshi.

Two important caveats:

1. I decided not to include much ground-fighting/jujitsu/BJJ in my regimen because out of the three major categories of combat sports (e.g., striking, grappling, and ground-fighting), ground-fighting is, at least for me, the one that lends itself the least to the kind of mental imagery training I was developing. There’s something about having your centre of mass so close to the ground and using your opponent’s body for extra leverage to overcome the confinement of this proximity that I simply didn’t feel I could adequately recreate all the underlying kinetic dynamics alone. So, I left it out of my program for the most part.

2. While wrestling techniques are undeniably effective in taking down your opponent, they typically entail going down to the ground yourself and ending up in a dominant and controlling position—ideally. Since I opted out of ground-fighting, I mostly favoured judo techniques—and one aikido technique—in which you remain standing after flipping your opponent to the ground. Also, my knowledge of wrestling is rudimentary, even though I’ve landed the odd double-leg takedown here and there during combat in the past. So, it only made sense not to include in my mental imagery program technical skills I hadn’t even fully “downloaded” via physical training.

With the fundamentals covered, I decided to sprinkle in some additional techniques that I happen to like. You see, while the basics represent techniques with the highest success rates regardless of underlying factors like age, sex, weight, height, reach, etc., interindividual variability must also be considered to ascertain what techniques, outside of fundamentals, may work best for you. And for me, God knows I love me some sidekick! So I added to the mix some elbows, some dynamic trips/takedowns, and of course, Bruce Lee’s battering ram of a sidekick. These fundamentals would, thus, make up the core of the techniques I’d be drilling during simulated sparring bouts.

So far so good, but what about the actual mental imagery protocol? According to the scientific literature, the so-called mixed package is what gives you the biggest bang for your buck. But since I’m not too big on motivational self-talk—it feels disingenuous to me—yet very big on keen observation, I decided to use video feedback as a complement to visualization, the efficacy of which was investigated in a new systematic review by Yazgan et al. (2026) that looked at 9 publications with 339 total participants. Defined in the paper as active observation (AO), the act of watching a video of someone performing a movement meant to be mimicked by the viewer may actually elicit mirror neuron activity in the motor cortex, with even greater response when imagining the associated physiological sensations.

The picture of my mental imagery training module was becoming clearer in my mind’s eye, as it were. I’d do some simulated sparring against an imaginary opponent while recording myself, rewatch the video while resting, visualize myself going over the same techniques while trying to feel the sensations involved, and repeat the process N number of times until the cumulative visualization period approximated the science-based “best practices” recommendation of 10 minutes per session of mental imagery.

Ok, Monday, Wednesday, and Friday (or Saturday): check! What about Tuesday and Thursday, the cardiovascular module? Due to logistics, I opted for a modified version of the RST. You see, nothing gets my heart racing more than sprinting up the staircase, and since I have 24/7 access to the stairway in my building, my choice was pretty straightforward. I simply asked ChatGPT to come up with a good substitute for the reference RST program where I’d be sprinting up the stairs instead of the garden-variety linear dash. Here’s the transcript of my prompt and the reply I was given:

Don’t get bogged down by the naming convention used by ChatGPT when referring to “reps” and “sets.” I’ve been using a different nomenclature so far, where its “3 sets × (6 × 6–10 seconds hard uphill stair sprints)” would translate into my “3 cycles of 6 reps of 6-10 seconds…” And so, I decided on the following RST protocol—which I suitably named staircase RST (s-RST): 3 cycles of 6 reps of 8 seconds of hard uphill sprints, with 30-second inter-rep and 4-minute inter-cycle rest/recovery periods. And that takes care of Tuesday and Thursday, completing 5 weekly sessions of interleaved routines that alternate between a mental imagery and a cardiovascular module. The following sections will take you through my entire program step-by-step.

MENTAL IMAGERY MODULE: 3 SESSIONS/WEEK

Whether I’m performing the mental imagery or s-RST module, I start each session with some Hung Gar Kung fu breathing katas, specifically chi kuen yat, chi kuen yi, and chi kuen sam—which loosely translate to energy fist one, …two, and …three, respectively, but they’re known to us at my martial arts academy as part one/two/three kuen (see video below in which I perform chi kuen yat). I once saw a video of the legendary Dutch-American mixed martial artist (retired), Bas Rutten, explaining the importance of stretching out and expanding your respiratory muscles (e.g., diaphragm & internal/external intercostals) before each training session to help promote seamless breathing. Breath is everything, breath is life—and if you can “lube up” your breathing apparatus before doing something “cardiovascularly” taxing, then you can squeeze out the most out of your endurance and stamina. While Bas has his own routine, for me, breathing katas can’t be beat, not only because they stretch and expand my respiratory muscles, but also because they really help me zone in and prepare mentally for the session ahead.

After the breathing katas, I start the two phases of the module. Phase I is comprised of 3 sets of simulated sparring, video feedback, and resting visualization. Here’s the breakdown:

• Simulated sparring (60s-120s): Before the bout, I choose a handful of high success techniques as well as other ones I particularly like and/or feel comfortable doing. Once I press “record” on my mobile device, I commence the bout, chaining the selected technical elements together as I fight an imaginary opponent. My focus here is not so much on the mental game—although I do try and visualize my opponent—but actual physicality including footwork, movement, and technical delivery (see video below).

• Video feedback (60s-120s): I sit in a comfortable position and watch the video of the previously recorded bout. I mentally catalogue things I like and dislike in the service of promoting the former and demoting the latter in future bouts.

• Resting visualization (~60s): In the same seated position as before, I close my eyes and rehearse the bout in my mind, and where possible, try and rectify in my imagination the things I disliked during video feedback while repeating the things I liked.

After my third and final set of the above components, I begin Phase II, which consists of 3 sets of “augmented” simulated sparring and resting visualization. Here’s the breakdown:

• Augmented simulated sparring (60s-120s): With my eyes closed, I spar my imaginary opponent anew. This time, I crank up the mental game by really trying to “see through the mind’s eye” as I attempt to implement what I’ve absorbed during the previous phase (see video below in which I primarily focus on straight punches and front kicks, with my eyes closed).

• Resting visualization (~30): Since I’m already integrating visualization in the simulated bouts, this component is typically shorter than its Phase I counterpart. I mostly use it for recovery and mental preparation for the next set.

NB: in keeping with the PETTLEP framework, I’d ideally be wearing my sparring equipment (e.g., mouthpiece, gloves, shin pads, etc.), which I may start doing down the road. But do notice how I always bow before and after the simulated bouts. That’s one way I recreate the physical dimension of the actual contest, that is, by observing the same procedural etiquette and rules of conduct I’d abide by when sparring real opponents at my martial arts academy.

When the two phases of my mental imagery module are complete, I typically finish off the session with about 20 minutes of free/body weight exercises (e.g., dumbbell presses, biceps curls, shoulder flies, push-ups, pull-ups, & sit-ups).

A SMALL DIGRESSION ABOUT KATA

In addition to the previously mentioned breathing forms, I like to do extra kata practice before my s-RST sessions. The combined Sil Lum/Hung Gar style of Kung fu I practice encompasses so many beautiful katas—including the 5 animals (e.g., tiger, crane, snake, leopard, & dragon)—it’d be a shame to restrict myself to only the 3 breathing ones (see video below where I perform Sil Lum Kuen Yat). Until about a few years ago, I must admit I had a love-hate relationship with kata. Part of me was very fond of it because I’m genuinely graceful and technical in my delivery, while the other part used to hate it because doing kata alone will never make you a good fighter. Then one day, Sifu said something that changed everything: “Don’t look at kata with realistic eyes.” As far as combat is concerned, kata is never meant to replace it, but only to complement it. Kata is a training paradigm in which the practitioner’s ultimate objective is to unify mind, body, and breath into one. When these three elements merge, you become Bruce Lee’s well known water that “becomes the cup!” I recall Joe Rogan talking about how his Yoga practice enhances his BJJ game. Of course, training Yoga alone will never enable you to slap on a triangle choke or slip in an Anaconda, but for Joe and other BJJ practitioners, Yoga somehow complements their ground-fighting finesse. To me, since I no longer look at kata with realistic eyes, it now serves a similar purpose.

CARDIOVASCULAR MODULE: 2 SESSIONS/WEEK

After doing my katas, I go to the nearest stairway and crank out the following: 3 cycles of 6 sets of 8-second “sprint-ups” and 30-second “walk-downs” with 3-min-and-30-second inter-cycle rests (see Fig. 7). To keep this aspect of my overall program challenging, every week that goes by, I try and add 1 second to the high intensity sprint-ups while subtracting it from the low intensity walk-downs. So far, the first cycle has felt like a breeze to me, and the compounded 4 minutes of rest towards its end feels like overkill. But after the second cycle, and especially the third one, I usually feel like I’ve maxed out.

And now, ladies and gentlemen, you’ve been fully debriefed on my solo training program.

PUTTING INTO CONCRETE FORM THE FINAL PRODUCT OF MY BRAIN

Like a novelist who hesitates to put pen to paper, the ideas and concepts that Tesla hatched up would’ve been no good if they hadn’t eventually left the coop of his mind: the alternating current (AC) system, hydroelectric power plants, the radio, and even the remote control may never have seen the light of day if Tesla hadn’t, as he said it himself, “put into concrete form [the] final product of [his] brain.” The same is true of Kashiwazaki’s solo training, which included mental imagery, the effectiveness of which would’ve remained unknown had he not competed in judo tournaments and won them. History is full of remarkable human beings whose propensity to effortlessly chart the landscapes of their minds offered them a unique advantage over their contemporaries within their individual disciplines.

Leonhard Euler, the 18^th century Swiss polymath—and arguably the most prolific mathematician of all time—was known for his incredible mental faculty to conduct lengthy algebraic computations entirely in his mind. Astonishingly, Euler had gone almost completely blind in both eyes by 1766, but instead of slowing down, his aptitude for complex mental calculations enabled him to increase his productivity, oftentimes dictating entire manuscripts to scribes. In 1775, it is believed that his academic output was on the order of one mathematical paper per week, with about half of his nearly 900 treatises produced after becoming blind, many of which were published posthumously.

And what of Fyodor Dostoevsky, who very likely suffered from temporal lobe epilepsy? Morgan (1990) stated the following in relation to his ecstatic auras:

Dostoevsky said that for several short moments before a fit he felt a contentment that was unthinkable for those who had not experienced it. He felt in perfect harmony with himself and with the entire universe. The perception was so clear and the episode so pleasurable that he expressed that he would give 10 years of his life and perhaps all of it for a few seconds of such bliss […]

These subjective episodes of mental clarity and perceptual acuity are believed to have heavily inspired the Russian wordsmith in creating some of the most celebrated fictional characters and literary works in human history, including Prince Myshkin from The Idiot who, like Dostoevsky, also suffered from epilepsy.

All this to say that I’m completely aware that all the work I’ll be putting into my 12-week training program will ultimately amount to nothing more than a temporary “side quest” if I don’t eventually put it into concrete form. To do so, I’ll have to go back to my martial arts academy in Montreal after the training period ends and engage in actual combat with my juniors and seniors. The kwoon (e.g., dojo in Japanese) will host the final stage of my experiment. Without it, I’m no better than the novelist who never put pen to paper. Until then, I’ll be adhering to my program as religiously as I can, adjusting it as required while monitoring my progress.

At the time of writing, I’ve completed nearly two weeks of my self-made, science-based combat regimen. While it’s too early to tell, my subjective assessment so far is that my visualization and technical execution—not to mention my cardiovascular fitness—all seem to be improving. I’m thinking of incorporating some kind of mid-program (e.g., post-6 weeks) appraisal, like a shuttle run test, that I’ll be repeating at the end of the 12 weeks. Ideally, I should’ve taken baseline measurements prior to beginning the program, but I was too excited to start, so I jumped the gun a bit.

It’s all good! My experiment is, after all, one with a very small sample size of N = 1, so I won’t be extracting any meaningful population-level statistical regularities from it anytime soon. Be that as it may, the prospect of putting myself through it, with its potential outcomes, has given me incredible motivation. As I told Sifu before embarking on this journey, “I haven’t been this excited about practice since I was a white belt!” I was a white belt when I started martial arts at the age of 10—I’m 41 now. Regardless of what comes of it, whether my program flops or passes with flying colours, there will be—I’m sure of it—important insights that I’ll have gained throughout the process. One way or the other, I’ll be sure to pass them onto my fellow Kujutsukas when I go back to Montreal for my ultimate test. Until then, stay tuned to ZAGROSCIENCE for updates!

Subscribe now

Share

Leave a comment

MY ANNOTATED ARTICLES

REFERENCES

1. Katsuhiko Kashiwazaki. (n.d.) In Wikipedia. Retrieved July 12, 2026, from https://en.wikipedia.org/wiki/Katsuhiko_Kashiwazaki

2. Tesla, N. (2021). My Inventions. Arcturus Publishing.

3. Pearson, J., Naselaris, T., Holmes, E. A., & Kosslyn, S. M. (2015). Mental imagery: functional mechanisms and clinical applications. Trends in cognitive sciences, 19(10), 590-602.

4. Van Caenegem, E. E., Moreno-Verdú, M., Waltzing, B. M., Hamoline, G., Lennart, F., & Hardwick, R. M. (2026). Greater neural overlap between motor imagery and working memory than with movement execution: A meta-analytic comparison. Imaging Neuroscience, 4, IMAG-a.

5. Slimani, M., Taylor, L., Baker, J. S., Elleuch, A., Ayedi, F. M., Chamari, K., & Chéour, F. (2017). Effects of mental training on muscular force, hormonal and physiological changes in kickboxers. J Sports Med Phys Fitness, 57(7), 1069-1079.

6. Andreato, L. V., dos Santos, M. G., & Andrade, A. (2022). What do we know about the effects of mental training applied to combat sports? A systematic review. Psychology of Sport and Exercise, 63, 102267.

7. Liu, Y., Zhao, S., Zhang, X., Zhang, X., Liang, T., & Ning, Z. (2025). The Effects of imagery practice on athletes’ performance: A multilevel meta-analysis with systematic review. Behavioral Sciences, 15(5), 685.

8. Thurlow, F., Huynh, M., Townshend, A., McLaren, S. J., James, L. P., Taylor, J. M., ... & Weakley, J. (2024). The effects of repeated-sprint training on physical fitness and physiological adaptation in athletes: a systematic review and meta-analysis. Sports Medicine, 54(4), 953-974.

9. Yang, Q., Wang, J., & Guan, D. (2025). Comparison of different interval training methods on athletes’ oxygen uptake: a systematic review with pairwise and network meta-analysis. BMC Sports Science, Medicine and Rehabilitation, 17(1), 156.

10. Yazgan, E. A., Kara Kaya, B., Tiryaki, P., Atlı, E., & Cavlak, U. (2026). Effectiveness of motor imagery and action observation in parameters of sport performance: A systematic review. Perceptual and Motor Skills, 133(1), 224-249.

11. Leonhard Euler. (n.d.). In Wikipedia. Retrieved July 18, 2026, from https://en.wikipedia.org/wiki/Leonhard_Euler

12. Morgan, H. (1990). Dostoevsky's epilepsy: a case report and comparison. Surgical Neurology, 33(6), 413-416.

In a previous essay1 on the topic of online neuroscientific charlatanism, titled Substack offered me a Faustian bargain: “As a neuroscientist,” I turned it down..., I discussed just how important contextual nuance is to truth as it pertains to scientific claims. To judiciously arbitrate its veracity, a proposition must be set within the confines of well-established parameters, and this is true not only of scientific claims but any claim in general: context is everything. Consider the following passage in which I break down one such ill-defined assertion made by a self-anointed neuroscientist (highlight added by myself):

The vaguer a claim is, the more it tends towards epistemological void—asymptotic as that trajectory might be. In recent years, I have noticed one particular neuroscientific subject matter that I have spent some time studying integrate into the zeitgeist, that is, the default mode network (DMN). With numerous claims abounding (e.g., news, entertainment, social media, etc.), describing the DMN as the “ego” centre of the brain, the “inner narrator”, the “self-consuming loop,” or other some such characterizations, separating fact from fiction is becoming increasingly mystifying. Since the science of the DMN, itself, is far from “settled,” as with almost anything neuroscience-related, there is bound to be some modicum of sensationalism—if not, downright fake news—whenever the topic comes up in the wilderness of online information. In this piece, I provide some timely scientific background to adequately contextualize all the hype. While my objective here is not to validate nor “debunk” any specific proposition about what the DMN is or does, I do want to offer you, the reader, a general account of pertinent scientific findings so as to equip you with some of the means by which you can parse matters for yourself.

SCIENTIFIC OVERVIEW

Biswal et al. (1995) is the study that set the field in motion, laying the groundwork for the discovery of whole-brain functional networks, including the DMN. Researchers wanted to know if there is an underlying organization to spontaneous low-frequencey fluctuations (<0.1 Hz) observable in the resting human brain via functional MRI (fMRI). They scanned eleven healthy individuals while they performed a bilateral finger-tapping task to identify regions of the sensorimotor cortex. Resting-state time-series were additionally acquired and used to compute correlation coefficients within these areas. Authors reported a “high degree of temporal correlation” within and across hemispheres in regions of the brain associated with motor function (Figure 1). This paper introduced the paradigm of functional connectivity, demonstrating that naturally-occurring resting-state brain fluctuations are structured and coherent, while paving the way for future studies to uncover the organization of the DMN using the same method.

Raichle et al. (2001) explored the idea that task-induced cortical deactivations might represent a deviation from a physiological baseline state in the adult human brain. Titled, A default mode of brain function, the study used positron-emission tomography (PET) to measure the oxygen extraction fraction (OEF), which is the ratio between oxygen used by and delivered to the brain. Metabolic baseline measurements from three independent groups of healthy participants were compared to meta-analytical data derived from nine previous neuroimaging studies (Figure 2). The paper noted remarkable OEF uniformity throughout the brain at rest, indicative of a baseline metabolic equilibrium. The study also found that, once a person engages in goal-oriented tasks, deactivations from this established baseline occur within specific midline regions, namely, the posterior cingulate cortex (PCC), precuneus, and medial prefrontal cortex (mPFC) (Figure 3). These findings hinted at an underlying network of distributed, yet related cortical structures involved in the brain’s functional “zero set point.”

The first direct evidence of the DMN came from Greicius et al. (2003) in which three questions were asked:

1. Is there a collection of brain regions that reliably shows increased activity at rest compared to objective-oriented cognition?

2. Is this network affected by simple sensory processing?

3. How is this system modulated by mental tasks?

Researchers used fMRI to scan fourteen healthy young adults under three conditions:

1. working memory task

2. passive viewing of a visual stimulus

3. resting-state with eyes closed

The PCC and ventral anterior cingulate cortex (vACC), which deactivated during working memory, were used to seed subsequent functional connectivity analyses in the resting-state and passive viewing conditions. Resultant cortical connectivity maps for both paradigms were “virtually identical,” showing significant coupling between the PCC and vACC as well as other areas, suggesting minimal modulation of this network by simple sensory processing (Figure 4). Interestingly, lateral prefrontal regions that were activated during task showed inverse correlations with the PCC at rest, indicating the existence of a dynamic mechanism by which the baseline “default mode network” is attenuated during goal-driven cognitive processes (Figure 5).

In a seminal paper by Yeo et al. (2011), a clustering algorithm was applied to the resting-state functional connectivity data of 1,000 subjects, yielding four stable parcellation schemes, of which, the 7-network solution is commonly referred to as the canonical resting-state networks (Figure 6, left). These networks include the Visual, Somatomotor, Dorsal Attention, Ventral Attention—or Salience—Limbic, Frontoparietal, and—of course—Default. (Figure 6, right).

In more recent years, with the advent of sophisticated machine learning techniques, connectomes (e.g., connectivity matrices) have been used to derive cortical gradients, capturing smooth topographic transitions in brain properties along the cortical mantle. In the age of Big Data, these approaches have expedited neuroimaging analytics by enabling researchers to decrease data complexity with no considerable loss to information content. In a process generally referred to as dimensionality reduction, “manifolds” can be derived from a starting dataset, corresponding to streamlined components that are stratified according to how much variance each one explains. In neuroscience parlance, these manifolds (or components) are the aforementioned cortical gradients. Margulies et al. (2016) used a similar technique on resting-state functional connectivity data to describe principal gradients of brain organization, notably demonstrating that the DMN occupies an apex locus that is structurally and functionally farthest away from “unimodal” areas involved in vision, audition, somatosensory processing, and movement (Figure 7).

In a review written by my friends and colleagues, Smallwood et al. (2021) expatiated on the complex forms of mental processes (e.g., episodic memory, social cognition, mental time travel, etc.) supported by the DMN, framing its topographical distance from primary sensorimotor systems as an explanatory paradigm. The review also noted that, while it is not part of the so-called “multiple demand” system tuned to “attention-demanding tasks” that exhibit “task-positive” behaviour, the DMN is nonetheless heavily involved in external goal-directed cognition when decisions must be guided by prior experience, internal rules, or schema-based information, speaking to its multifaceted functionality.

What we have seen of the scientific literature so far has primarily addressed the DMN in the context of function, which is ultimately constrained by physical structure. Thus, to fill a glaring lacuna pertaining to a fine-grained anatomical account of the functional multiplicity of the DMN, in Paquola et al. (2025), my colleagues and I brought to light the significance of its microarchitecture. Specifically, we combined postmortem 3D histology with in vivo multimodal neuroimaging (e.g., diffusion MRI, quantitative T1 relaxometry, and resting-state fMRI) to determine how DMN cytoarchitecture contributes to its structural wiring and information flow. The cellular morphology of associated brain regions displayed heterogeneity, with characteristics consistent with both unimodal (e.g., visual & auditory processing) and heteromodal (memory & self-referential abstraction) computations (Figure 8).

Using diffusion map embedding, a nonlinear manifold learning approach, we derived a principal cytoarchitectural axis—first eigenvector or E1—that described variations in laminar differentiation and cellular density in the DMN. Regions with lower E1 values (e.g., retrosplenial & posterior middle temporal cortex) acted as “receivers,” gathering convergent input from sensory hierarchies. Conversely, areas with higher E1 values (e.g., parahippocampus & anterior cingulate cortex) represented an “insulated core” that is “decoupled from perception of the here and now” (Figure 9). By highlighting underlying cytoarchitectonics, our study was the first to describe the DMN as an “association hierarchy” where microstructural heterogeneity translates into functional polyvalence.

THE ROAD AHEAD

Over 30 years of rigorous scientific investigation have gone into defining and characterizing the DMN. Deactivated during attention-demanding tasks and involved in complex, internally-oriented cognition, this network consists of a set of functionally interconnected brain regions distributed across the cortical mantle, with a multilayered and heterogenous cytoarchitectural composition. While much progress has been made in the field since the mid-1990s and early 2000s, many questions remain unanswered. Among these pressing issues are:

1. the cellular mechanisms that support the dynamic balance between excitation and inhibition in the DMN

2. the characterization and functional relevance of sub-networks within the DMN

3. the functional integration of the DMN with other canonical networks for sustaining subjective continuity

4. the role of DMN alternation in psychopathology and how it may be targeted therapeutically

5. the development and progression of the network across lifespan and its evolutionary origins

The above is, by no means, an exhaustive list, but addressing them and other related questions in the years and decades to come will have important scientific implications. Breakthroughs will be precipitated by progress in neuroimaging analytics, computational modelling, causal interventions, and translational research. Advances in these areas will transform our understanding of the self, cognition, mental health, and consciousness, and I, for one, am excited to see what the field has in store, awaiting for the next big blockbuster discovery, the science of which, I will be sure to share with you.

BIBLIOGRAPHY

Biswal, B., Zerrin Yetkin, F., Haughton, V. M., & Hyde, J. S. (1995). Functional connectivity in the motor cortex of resting human brain using echo‐planar MRI. Magnetic resonance in medicine, 34(4), 537-541.

Greicius, M. D., Krasnow, B., Reiss, A. L., & Menon, V. (2003). Functional connectivity in the resting brain: a network analysis of the default mode hypothesis. Proceedings of the national academy of sciences, 100(1), 253-258.

Margulies, D. S., Ghosh, S. S., Goulas, A., Falkiewicz, M., Huntenburg, J. M., Langs, G., ... & Smallwood, J. (2016). Situating the default-mode network along a principal gradient of macroscale cortical organization. Proceedings of the National Academy of Sciences, 113(44), 12574-12579.

Paquola, C., Garber, M., Frässle, S., Royer, J., Zhou, Y., Tavakol, S.2, ... & Bernhardt, B. C. (2025). The architecture of the human default mode network explored through cytoarchitecture, wiring and signal flow. Nature neuroscience, 28(3), 654-664.

Raichle, M. E., MacLeod, A. M., Snyder, A. Z., Powers, W. J., Gusnard, D. A., & Shulman, G. L. (2001). A default mode of brain function. Proceedings of the national academy of sciences, 98(2), 676-682.

Smallwood, J., Bernhardt, B. C., Leech, R., Bzdok, D., Jefferies, E., & Margulies, D. S. (2021). The default mode network in cognition: a topographical perspective. Nature reviews neuroscience, 22(8), 503-513.

Yeo, B. T., Krienen, F. M., Sepulcre, J., Sabuncu, M. R., Lashkari, D., Hollinshead, M., ... & Buckner, R. L. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of neurophysiology.

Dedicated to my People and the Fire that forever burns in us.

DISCLAIMER

This story is a work of fiction. Names, characters, places, and incidents are either the product of the author’s imagination or are used fictitiously. Any resemblance to actual events or persons, living or dead, is entirely coincidental. And yet…

GETTING SETTLED

You’re an interstellar researcher visiting an alien planet to learn about its denizens. Called Heart, this planet is eerily similar to your own, from its geography down to its people. Specifically, your assignment is to study the citizens of Rain, a country on planet Heart where the Serpians comprise the largest ethnicity and which also includes a sizeable Durkish minority. Unfortunately, you’re restricted from visiting Rain due to ongoing geopolitical and civil unrest, obliging you instead to conduct your research from another country on a different continent. Even so, you have unrestricted access to the most reliable textbooks and up-to-date information databases on the planet.

One tidbit you pick up early on is that the toponym “Rain” is oftentimes used interchangeably with “Serpia,” which designates the landmass that was once part of a bygone Serpian empire (~2,500 years ago). Curiously, you also learn that the term “Rainian,” which technically encompasses all citizens and ethnic groups of Rain, is almost always used synonymously with “Serpian.” For example, the expressions “Rainians are a noble people” and “Serpians are a noble people” are, to all intents and purposes, one and the same. Incidentally—and to make matters even more confusing and ambiguous—you find out that the Serpian and Durkish languages are subsumed under a broader linguistic category also known as “Rainian” (or “Rainic” in more academic circles). Oh, and “Serpian” is also known as “Sarfi” by the way… Yeah, it’s a lot to take in all at once.

So, you write yourself a memory aid to make sure you have all the basics down:

1. Rain is a nation state on planet Heart.

2. “Rain” and “Serpia” (e.g., the territory controlled by the ancient Serpian empire) are typically used synonymously.

3. A Rainian is someone who lives in Rain and/or someone who speaks a “Rainian” (or “Rainic”) language, such as Serpian/Sarfi or Durkish.

4. “Rainian” is oftentimes used interchangeably with “Serpian” (e.g., the ethnic majority in Rain), but not with “Durkish,” even though Durks, just like Serpians, live in Rain and speak a Rainian/Rainic language.

Ok, so far so good!

THE DURKISH QUESTION

Just when you think you’re starting to get the gist, you make a shocking discovery about the Durks: most of them are not Rainian—at least not by citizenship. In fact, world population statistics reveal that fewer than one-third of all Durks on planet Heart live inside Rain, leaving a majority of them uniformly spread across a handful of contiguous and neighboring states. A survey of these other countries shows that most Durks, with the exception of maybe those living inside Rain (e.g., Rainian Durks), don’t actually self-identify as Rainian. You find yourself in a pickle! You see, your mission is to study and write a repot on the people of Rain—so how can you make any sense of the Durkish minority when most Durks live beyond the country’s borders? Looking at the maps of these other nations together with that of Rain, it almost seems like their shared borders are breaking apart what otherwise appears to be a single Durkish homeland, marked by mountainous landscapes, which goes by the name of “Durkistan.” Your sources, which include genetic studies, also indicate that the Durks are indigenous to Durkistan, specifically, the highlands of Sargoz.

You decide to do more research on Durkistan. After an entire week of immersing yourself in learning about its history, climate, people, and geography, you make several astonishing discoveries:

1. Agriculture first emerged in Durkistan (12,000–10,000 years ago).

2. The oldest megalithic structures were erected in Durkistan (~12,000 years ago).

3. Clay tokens—precursors of written language—were initially used by highlanders in Durkistan (11,000–10,000 years ago).

4. The earliest domesticated farm animals were husbanded in Durkistan (11,000–10,000 years ago).

5. All major Heartling languages trace back to a geographical epicenter in Durkistan (~8,000 years ago).

6. Beer made from barley was first brewed in Durkistan (~5,500 years ago).

7. The oldest surviving piece of musical notation was composed in Durkistan (~3,500 years ago).

8. The first extant treatise on how to train chariot horses was written in Durkistan (~3,500 years ago).

And the list goes on…

Based on your findings, you determine that Durkistan is a serious contender for the cradle of Heartling civilization. While your conclusion is borne out by a rigorous synthesis of available data, you find it bizarre that nearly all contemporary Heartling scholars refrain from using terms like “Durkistan” and “Durks” in their works when discussing the geographical locations of the aforementioned foundational milestones and their native inhabitants. Instead, they seem to favor a contextualization scheme predicated on the geopolitical boundaries of modern nation states, such as Rain and its surrounding countries, setting arbitrary constraints that fragment what is otherwise a unified and continuous series of civilizational phenomena. Whatever the reason for this glaring systematic omission, its net effect is to cast an opaque veil on Durkish history, prompting you to remark, “It sucks to be a Durk! Imagine being indigenous to the birthplace of civilization only to be forgotten while you still exist.”

ARE DURKS REALLY RAINIAN?

You remind yourself that, as fascinating as Durkistan is, your objective is to study the country of Rain and its population, which, while including a substantial Durkish minority, is, first and foremost, a Serpian nation. Indeed, the official language of the country—from government and education to media and entertainment—is Serpian/Sarfi. On a side note, since the religion of Samil is such a huge aspect of the spiritual fabric of Rain, it should also be noted that its ruling clerical elite, consisting primarily of Serpian clergymen, in addition to using Serpian/Sarfi during Friday sermons, also resorts to Abaric as needed, for it is the exclusive language in which Samilic holy scriptures are written. After some deliberation, you turn your attention to Rainian/Rainic languages, that is, Serpian/Sarfi and Durkish. At first glance, they seem to overlap significantly as far as vocabulary goes, but then you make another startling discovery: nearly half of the Serpian/Sarfi lexicon consists of Abaric loanwords. But don’t ever tell that to Serpians, because they absolutely hate it when you draw comparisons between them and those “savage Abars,” an expression used by some Serpians to vent their disdain toward Abaric people, even though Serpians have historically embraced both their religion (e.g., Samil) as well as their language. Surely the same must be true of the Durks who are, just like the Serpians, linguistically Rainian/Rainic. Well…no, not really! In fact, you are able to determine that:

1. Abaric loanwords in Durkish are only on the order of ~10%.

2. For almost any Abaric loanword in Serpian you can identify, the Durkish language has its own distinct terms—that’s right, not just a single unique word, but multiple synonyms.

According to some of your sources—although much more academic work is needed for resolution—Durkish could possibly be the richest and most diverse language on Heart when synthesizing all its dialects and regional variations. It’s almost as if the precipitous landscape of Durkistan had shielded Durkish from too much external influence, allowing it to retain an archaic character that many other languages had lost to time. These considerations prompt you to question the very validity of its categorization as a Rainian/Rainic language. “It’s like cramming a size-10 foot inside a size-7 shoe…” you tell yourself, “…and pretending that it fits.” As for religion, while a majority of Durks do, indeed, practice Samil, they are overall much more diversified in their doctrinal worship than Serpians. As a matter of fact, Durks also practice a series of ancient faith-based creeds that are uniquely theirs, such as Elavism, Saryanism, and Zayidism.

Naturally, you begin to wonder whether or not Durks are, in fact, Rainian. The designation seems flat, inadequate, and somewhat arbitrary. Indeed, most Durks live outside of Rain, don’t self-identify as “Rainian,” are native to a continuous swathe of mountainous land known as “Durkistan” (e.g., the birthplace of Heartling civilization), speak a complex system of interrelated ancient dialects that shows a relatively low degree of linguistic admixture with external languages, and practice a variety of different religions—some of which are uniquely their own. And that’s not even considering geopolitics—Durks, as it turns out, have historically struggled for greater self-determination across the different nation states that converge over this ancestral homeland of theirs. In fact, another eye-opening discovery you make is that Durks are the largest nation on planet Heart without a country of their own, with a population of approximately 45 million, and possibly even more. Holy moly, that’s a lot of stateless people! How Rainian can they really be? They clearly aren’t Serpian, that’s for sure. So why use a term to label them that virtually always harkens back to Serpians? Is it fair? Is it academic? Is it true?

THE PUB INCIDENT - PART I

After two months of hard work learning and writing about the people of Rain, it’s time to wind down. Your Heartling friends and colleagues decide to take you out to a local pub. Among them is Arileza, a PhD candidate and Serpian exchange student from Rain, who’s always been kind and extremely generous with his time in assisting you with your research about his home country. After a few drinks that jazz up the mood, you insist on buying the next round as a token of your appreciation for all the hospitality you’ve experienced throughout your stay on Heart. So you walk over to the counter and wait for the bartender to finish serving a previous patron seated on the barstool, a certain lady whose striking necklace immediately catches your eye. “Durkistan!” You exclaim, as a smile spontaneously draws on her lips. “How did know?” She inquires whimsically. “I do research on the Durks,” you reply as you point to her necklace, “I knew right away that was the Durkish map when I saw it.”

The conversation that ensues is truly enlightening. You learn from your new acquaintance, Jorîn, that Durks have their own toponyms for the different parts of Durkistan: North is “Kabur,” South is “Şabur,” East is “Hojrelat,” and West is “Ojrava.” She, herself, is originally from Eastern Durkistan or Hojrelat—the closest thing you can think of is “Rainian Durkistan.” Jorîn tells you that, although all those expressions designate the same region (e.g., Eastern Durkistan, Hojrelat, and Rainian Durkistan), she’s not fond of the last one: “I don’t like using ‘Rainian’ or ‘Rain’ for anything Durkish related,” she says, “All these terms are good for is to bolster a false narrative by the Serpians that anything Durkish is somehow also theirs.” You’re genuinely intrigued, so you press her on: “What do you mean?”

- “Well,” she begins, “take our new year’s festival ‘Roznew,’ for instance. Every year on the vernal equinox, we Durks light a huge bonfire and dance around it, hand-in-hand, everyone dressed in colorful, traditional garbs. In some Durkish cities, we hike up mountains—hundreds of people in a single file, each one carrying a torch to the summit. Once at the top, one by one we throw our fires into a massive cauldron to light it up. This yearly ritual is a tribute to our national mythology, commemorating the victory of Waka the Blacksmith over Zahuk the Tyrant. Serpians, well, they don’t do any of that, but they sure like to take credit for it by falsely claiming that our celebration is ‘Rainian’ or ‘Serpian.’ To usher in their new year, they’ll instead party indoors over a ceremonial table called the ‘Faht-Seen’ or go outside to watch street performers clad in red and wearing ‘black-face’ playing the tambourine and singing about their ‘Lord.’”

- “What?” You’re legitimately confused.

- “I mean, I’m not judging or anything. More power to them. They can celebrate their new year however they choose, but don’t [expletive] call it ‘Rainian’ or ‘Serpian’ when we Durks do our own thing. The point is, if you want to have a real good time, you’re better off partying with the Durks!”

After some time, you come back to your group carrying a tray of shots, lost in thought and somewhat detached. “What took you so long?” One of your colleagues asks, but his question doesn’t even register with you. As more voices wrangle you in, you snap out of it, apologizing for your aloofness.

- “Everything ok?” Arileza inquires as he takes a seat next to you, “What was that all about?”

- “Have you ever had the feeling…” you respond, “…that no matter how much you’ve learned about certain people in textbooks and databases, there’s always something more to know about them when you see them in real life?”

- “Yeah…I mean, textbooks and databases can only account for so much of the total variance in what makes people, people. Sometimes, the best way to know someone is to just sit down and engage them in a conversation.”

- “Exactly! I just spoke to a Durkish girl at the bar who taught me so much about her people that I simply didn’t know before.”

- “Dorkish? From Rain?” [Some Serpians have a surprisingly hard time correctly saying Durk, Durkish, and Durkistan; instead they pronounce these as Dork, Dorkish, and Dorkestan, respectively.]

- “Umm…yeah, although she wasn’t particularly fond of…”

- “We love our Rainian Dorks! They’re such brave people. In many ways, they embody some of the best qualities about us Rainians.”

- “Yeah…she seemed very headstrong, and I mean that in a good way.”

- “Dorkish women are Rainian lionesses! They’re well-known for being fierce warriors. Speaking of which, did you know that the first female military commander in recorded history was someone from ancient Rain? Her name was Tawarinum, and she is said to have lead an army of over 10,000 soldiers some 3,800 years ago.”

- “3,800 years ago? But that’s way before any mention of ‘Rain’ or ‘Rainians.’”

- “Technically, yes, but the region is historically known as Rain or Serpia.”

- “Where exactly in Rain was this Tawarinum from?”

- “So, supposedly she hailed from the Sargoz mountains.”

- “Sargoz mountains, you say? Well, that’s Durkistan, isn’t it?”

- “Well, yes and no. First of all, as you already know, Dorkestan isn’t an actual political entity; it’s just a loosely defined geographical region with a Dorkish majority. Secondly, the name ‘Dorkestan’ didn’t even exist 3,800 years ago. So, it’s kind of anachronistic to retroactively label ancient regions using contemporary toponyms.”

- “Isn’t this exactly what you’re doing by using terms like ‘Rain’ and ‘Serpia,’ which also didn’t exist at the time? My question is: why use those names and not ‘Durkistan?’ Also, so what if it’s not a well-defined, modern political entity? Lots of geographical names are similar to ‘Durkistan’ in that sense, and we still use them—case in point, Atanolia and Mesotomapia.”

- “I mean, sure, but, speaking of textbooks and databases, these generally use expressions like ‘ancient Rain’ or ‘Serpia’ to designate the Sargoz mountains and places otherwise associated with Dorkestan. Besides, at the end of the day, whether Serpian or Dork, we are all Rainians.”

- “But don’t you owe it to the preservation of Durkish heritage to articulate what specifically relates to them with more precision? Take the 14^th century Serpian poet Zefah, for instance. There’s no ambiguity about his ethnicity. He’s known, the world over, as one of the great Serpian poets. And then, you have someone like the 9^th century polymath, Diwanari, who hailed from Durkistan and even wrote a book titled ‘Ancestry of the Durks.’ There’s so much debate about whether he was Durkish or Serpian, even though the odds of him being Serpian given where he’s from are slim to none. Had he come from, say, the Sarf province of Rain and written a book titled ‘Ancestry of the Serpians,’ no one would have any qualm about him being Serpian. Instead, his Kiwidepia page, for example, is full of ‘Rainian’ this and ‘Serpian’ that. And he’s just one historical figure among many others whose ancestries are supposedly ambiguous because they were very likely ethnic Durks. Labeling them as ‘Rainian’ and declaring ‘case closed’ is somewhat disingenuous if you ask me, because you and I know all too well that most people will just assume that ‘Rainian’ and ‘Serpian’ are the same. That’s how you erase an entire people, that is, by whitewashing the history of the entire region.”

- “Our Dorks, as far as I know, have no issues being designated as ‘Rainian.’ In fact, they are some of our proudest Rainians ever.”

- “Why do you keep saying ‘our’ every time you talk about Rainian Durks? It’s not like you own them or something. And for the record, the Rainian Durk lady whom I just saw said that she doesn’t consider herself Rainian.”

- “She’s probably some Dorkish ultranationalist or leftist idealogue—or one of those separatist Dorks from outside Rain meddling in our internal affairs. As far as I can tell, Rainian Dorks are more than happy to be Rainians, as they should be! We have one of the oldest cultures, don’t you know!”

- “Well, why don’t we ask her how she feels about it?”

THE PUB INCIDENT - PART II

You spot Jorîn in the crowd and raise your hand, motioning her to join you. After introductions, you debrief her on the conversation you’ve been having with Arileza and ask for her two cents.

- “First of all, that is correct,” Jorîn begins, “I am a Hojrelati Durk from Rainian-occupied Durkistan…”

- “…what? What do you mean, occupied?” Arileza can’t believe his ears.

- “Yes, occupied! We Durks have been living under one Serpian government after another for over a century, whether it be the Monarchist regime or the Samilic theocracy, and we’ve had enough. We want our own autonomy, we want our own right to self-determination, we want our own independence… We’re tired of being treated as second class citizens on our own native lands.”

- “But I don’t…understand,” Arileza stammers, “This is the first time I’ve heard a Rainian Dork speak like this. I have many Dorkish friends and they’re all, very proud Rainians.”

- “Hey! What did I just say? I’m not ‘Rainian,’ so don’t make insinuations. I’m from Eastern Durkistan, aka Hojrelat, aka Rainian-occupied Durkistan. And for your information, it’s pronounced ‘Durk’ with a ‘u,’ not an ‘o!’ I can’t speak for your so-called ‘Dorkish’ friends, but maybe you should ask them, if they’re so very proud to be Rainian, if they’re equally as happy not to be taught their own mother tongue at school, or be allowed to give their children ‘Durkish’ names in some provinces and local districts, or live in some of the poorest sectors in the country due to governmental oversight or negligence, or even be part of a national minority that makes up a majority of prisoners of conscience and those who are executed for political dissent.”

- “All…all I’m saying, with all respect—you know, it should be said that I love all people—is that our Dorks are…”

- “…‘our Durks?’ What the [expletive] does that even? Have you ever heard a Durk use the expression ‘our Serpians?’ What the [expletive] does that say about you?”

- “…no…please, I mean, you’re our sisters and brothers—Dork and Serpian, we are all cousins, one big family, one nation…”

- “Well, we’ve tried being your sisters and brothers and cousins... And where has that gotten us? Appropriation of our culture, heritage, language, and history so that Serpian usurpers can make up some bloated, fairytale national identity for themselves… You people always talk about unity, family, brotherhood, humanity, and all that good stuff, but these are all empty sentimental platitudes that practically amount to nothing more than sweet little old lies. No, I’m not your sister and you’re not my brother, but I do want to be your equal. Until then, you keep your black-face and I’ll keep my fire!”

- “What do you mean ‘black-face?’ We have Lamshu, Serdowfi, Zefah… Rycus the Great! What do you people have? You should be so lucky, you mountain dweller!”

- No! What you actually have is a superiority complex, and unfortunately for me, thanks to many of my brainwashed compatriots, mine is an inferiority complex—our respective peoples are a match made in heaven. Oh! And your dear Rycus the Great was an ethnic Durk!”

- “Yes, exactly! His mother was Dorkish. But unlike you, I take pride in our shared Rainian ancestry.”

- “You know what, forget black-face. Keep your stupid Rainian ancestry, just give me back my land.”

What on Heart did you unleash? That went downhill, really quick! You step in between them and encourage them to keep it civil, as tempers begin to fray. Jorîn leaves abruptly to find her other friends, and you’re left with Arileza and the rest of your crew to try and digest what just happened. “My goodness, did you see her?” Arileza is visibly flustered, and it isn’t just the ethanol! “You know, many of them are gypsies and uncivilized people from the mountains.” You are reminded that he is referring to the same mountains that birthed Heartling civilization. In the end, the fire had extinguished almost as soon as it had ignited. With cooler heads prevailing, the rest of the evening goes by rather pleasantly—and you never get to say goodbye to that feisty Durkish flame.

ONE LAST TOUCH BEFORE BED

Today, four months have come and gone since arriving on Heart, and you still think about that eventful night at the pub nearly two months ago. You keep replaying the scene in your mind, wondering who was right: Arileza, and his insistence that Durks are Rainian, or Jorîn, and her spirited defiance and claim to the contrary? You’ve already drafted your treatise on the nation of Rain and its multifaceted peoples and history. Your stay on planet Heart is coming to an end, and you decide to give your manuscript one final look before submission, prior to embarking on your interstellar voyage back home, the duration of which will be spent in suspended animation. There’s one section in particular that has been gnawing at you ever since you wrote it, the one with the heading “The Durks of Rain.” You read the first two sentences:

Durks comprise one of the largest ethnic minorities in Rain, possibly the largest, although reliable demographic statistics are notoriously scarce. They inhabit a region in the West/Northwest of the country marked by its unforgiving, albeit picturesque highlands.

You read the passage over several times. Something is missing. Something real, which very few other contemporary sources overtly recognize—but you’re different. Your academic and intellectual integrity depend on it. And so you add the following:

This region, the Durks themselves call “Hojrelat,” with contiguous parts beyond the political borders of Rain, the totality of which constitutes their ancestral homeland, Durkistan.

Your cryogenic chamber awaits.

If to the moment I should say:

Abide, you are so fair—

Put me in fetters on that day,

I wish to perish then, I swear.

~ Goethe (Faust)

MEA CULPA

Back in January, I hosted on my podcast, The Abstract, where we discussed the role of midfrontal theta oscillations in response-conflict processing. At the outset of the episode, I announced that I had been drafting an article pertaining to the spread of pseudo-(neuro)science on social media (see video below). I was hoping to publish my piece sometime in February, but my plans got significantly derailed… Early that month, through no fault of my own, I was hurled into a living nightmare, a Kafkaesque reality that is still looming about, which effectively upended some of my priorities in life. When the time is right, I will go public with my story to raise awareness about what my family and I have had to endure in the hopes that no one else suffers the same fate—which, given the nature of human psychology and bias, is, perhaps, wishful thinking on my behalf… I won’t elaborate any further for now, but rest assured that everything will be made clear in due time.

Compounding my personal problems is the ongoing geopolitical unrest that has recently intensified across my homeland of Kurdistan. My platform, ZAGROSCIENCE, is dedicated to exploring the interface between neuroscience and other disciplines, such as philosophy, literature, history, and mythology. I have deliberately veered clear of political commentary, because it tends to be combustible. A simple perusal of discussion threads on social media outlets about global affairs suffices to drive this point home, and while I don’t necessarily think it’s always a bad thing to engage others in a heated factional disagreement, I believe the inevitable vitriol that oftentimes follows serves little to no purpose in discussing the brain/mind. Even so, my intentional omission of politics on this platform should not be misconstrued as an admission of apoliticality. Like almost everyone else, I possess my own opinions on international and sectarian dynamics, and as a Kurd, I admit that recent developments in the Middle East have been weighing heavily on my mind.

I should clarify that my relative inactivity on Substack as of late is not solely due to grievances—personal or otherwise; we all have our own daily demons to deal with, and I’ve never been one to kowtow in the face of adversity. As it turns out, throughout these difficult times, most of my cognitive bandwidth has been taken up by a grand project that I started about two months ago, which, I’m hoping, will lay the groundwork for how social media is used in the future—and I’ll be elaborating more on this initiative as the year progresses. Social media, in fact, is at the heart of this article, wherein I examine the corrosive effect of its Faustian bargain on the intellectual and moral integrity of individuals who specifically write and talk about neuroscience.

INFERNAL PACT

So, what exactly is a Faustian bargain? For starters, Faust is widely recognized as Goethe’s magnum opus, a timeless play hailed as a crowning achievement in German literature. It tells the story of a cosmic wager between The Lord and Mephistopheles centered around the nature of the human soul. As they set their sights on Faust, the titular protagonist, a doctor with a brilliant albeit dissatisfied mind, The Lord agrees to allow Mephisto to attempt to corrupt him with worldly delights (e.g., knowledge, pleasure, power, etc.), reassured that a striving good man, no matter how much he may stray, will always find the path to righteousness. Speaking to Mephisto, The Lord says:

Enough—I grant that you may try to clasp him,

Withdraw this spirit from his primal source

And lead him down, if you can grasp him,

Upon your own abysmal course—

And stand abashed when you have to attest:

A good man in his darkling aspiration

Remembers the right road throughout his quest.

Mephisto then appears to Faust in his study dressed as a traveler and makes him an offer. He swears to serve Faust on Earth and try to satiate all his mortal aspirations, offering him unbridled worldly experiences, if Faust pledges to accord Mephisto his soul in the afterlife in the event that he reaches a moment of complete contentment, wishing it to linger forever:

If to the moment I should say: Abide, you are so fair—Put me in fetters on that day, I wish to perish then, I swear.

In today’s digital world, Mephisto appears to us in our phones, computers, and tablets. And just like he made the good doctor his target, he is inclined to single out those among us who are seekers and purveyors of knowledge. In the next paragraphs, I will give examples of some such individuals who have fallen under Mephisto’s spell—people with large online audiences who have accepted the Faustian bargain of bountiful clicks, likes, comments, posts, and restacks—not to mention money—in exchange for their intellectual and moral integrity.

LOST SOULS

Consider the following note taken from an academically credentialed individual who curates of one of the largest “neuroscience” platforms on Substack1:

Ah! Now you know where I got my title from, but I’m sure many of you already had a hunch…

Starting off the sentence with As a neuroscientist, which is a hallmark of the author’s social media shtick, is their way of indicating that they are speaking from a place of authority. They’re essentially saying, “Listen to me, you plebe, because I know more than you.” Otherwise, if not to signal their self-proclaimed intellectual superiority, it’s unclear whether they mean that one needs to be a neuroscientist to appreciate the effects of showering in the dark, or it is thanks to neuroscientific investigation that they’ve concluded that showering in the dark can lead to resetting the nervous system. Also, what is actually meant by to reset my nervous system? Presumably, the author is saying that you should shower in the dark, because you’ll feel good afterwards. But how is that the same as resetting your nervous system, whatever that means? Indeed, the wording is vague, lazy, and disingenuous, offering no scientific context whatsoever to support his claims, other than, “Trust me, bruh!” The entire note, which masquerades as neuroscientific insight, basically boils down to an appeal to authority mixed with pseudoscientific fluff. Here’s how I’d rewrite it:

Behold, feeble-minded peasantry! Heed my words for I am thy purveyor of truth, and I hereby decree, “Fluff-a-dee-doo-dah!” and ye shall obey!

This person’s Substack page is a well of opportunities for this kind of dissection, so let’s do another one.

I love the above note, not because it offers a Neuroscience Fact—it doesn’t—but because it’s a prime example of the mind-numbing and intellectually corrosive effect of Substack’s Faustian bargain. Let’s break it down.

The brain’s main goal isn’t happiness. It’s safety.

Your brain—with its 65-85 billion neurons and 40-50 billion glial cells—is an intricate constellation of living tissue that performs highly elaborate operations, such as regulating autonomic functions (e.g., heart rate, blood pressure, breathing, etc.), controlling movement and coordination, processing sensory information, managing emotions and basic drives, and enabling higher cognitive actions. To suggest that it serves a singular main goal—whether it’s happiness or safety—grossly oversimplifies the functional complexity of the human cortex, and no self-respecting neuroscientist that I know would ever allow themself to publish such an intellectually lazy and sloppy statement.

That’s why you stay in situations that are slowly draining you. Familiar pain feels less risky than unknown possibility.

This segment builds on the false premise that your brain has a main goal and that this main goal is safety—whatever that means. Even assuming that the presupposition is correct, we’re still left with a general statement about how familiarity breeds decay and stagnation even if it feels safe. How this kind of cheap psychological triteness qualifies as a neuroscientific fact is beyond me.

Growth requires convincing your brain that uncertainty isn’t the same as danger.

I always cringe at language implying the existence of an agent that is independent of the brain that can somehow effect change upon it, such as convincing it to act one or the other way. That which we take to be the hub of our innate agency, from a neuroscientific perspective, is yet another layer of what the brain does, which is to say the construction of the ego or the sense of self. As a neuroscientist—which is what the author insists they are—making such an implicit distinction between the ego and the inner workings of the brain that give rise to it betrays either a fundamental lack of understanding of the underlying science or a disregard of it. Otherwise, the sentence articulates the idea that personal growth requires some measure of risk taking, which is basic wisdom. Good sense and sound judgment are free cognitive commodities that you should already possess—no fancy degrees in neuroscience needed—so be wary of professional swindlers who are trying to sell them to you.

Let’s turn our attention to another premium “neuroscience” platform on Substack and see what other golden nuggets we can excavate.

I’m going to need a sturdy pickaxe for this one…

There’s something that only your brain can do.

Enlighten me…

Figure out your thing, and go all in on being yourself.

Your thing? Go all in on being yourself? Woolliness passing for profundity—this is what I make of it; the equivalence of saying something without actually saying anything, akin to, “no one is as good as I am at being myself.”

No one can beat you at at being you.

No 💩, Sherlock!

Overall, the kind of redundant, vague, and circular tripe expressed in the above note is what I would describe as manifested mental constipation. It almost feels like the author—who has a PhD in cognitive neuroscience, mind you—is trying to push something big out…and then nothing comes out, except forced effluvium!

Here’s another example from the same account:

For starters, there is very little actual neuroscience in this note, and whatever is there is ambiguous.

When you come across information that differs from what you think is true your brain responds as if it were a threat.

While there is a modicum of truth to this statement, the wording lacks clarity and context. Your brain constantly compares information it receives from the senses to pre-catalogued models of how the world is supposed to be. When a prediction error occurs (e.g., a mismatch between expectation and observation), the cognitive model generally updates to accommodate the new information, sustaining flexible behaviour and learning. This process does not always trigger fear—if it did, we would be constantly aghast. Prediction error and model updating represent continuous and complementary features of the inner workings of the brain, spanning multiple tiers of information processing (e.g., low-level perception to higher order executive functions). In fact, even when it comes to information that contradicts an opinion—which is really what the note is about—unless said opinion is deeply rooted within one’s sense of identity, it is not necessarily the case that your brain will set off the alarm. I have written at length about this very topic, so if you’re interested in the neuroscience/psychology of belief formation and revision, I invite you to read my post, Harvest of Reality: The Cortical Organization of Belief.

Your brain sends the same signals as if your life were in danger, jolting you into fight or flight mode and shutting down the prefrontal cortex (rational, reasoning and judgment center).

How theatrical! Nuance is everything, and without it, it’s impossible to bear out the truth—but who cares about the truth when you can sound sexy, right? Does your sympathetic nervous system go into hyperdrive when your life is in imminent danger? Yes, absolutely. Is this reaction the same kind of fight or flight response as when, say, someone disabuses you of the false notion that the capital of Australia is Sydney? No! Nuance, people…nuance! Also, no—your prefrontal cortex (PFC) does not shut down either. Imagine being chased by a hungry lion in the African savanna. Fight or flight? You bet! Now imagine what would happen if your PFC actually shut down. Let me paint you a clear picture. Gripped with abject terror, the weight of inertia would come crashing down on you during your fleeting moments of life as the giant feline pounces on you, transpiercing your throat with its long, dagger-like canines. You see, since the PFC is heavily involved in decision-making and planning, you wouldn’t be able to choose what course of action to take (e.g., run, hide, climb a tree, etc.), so you’d be frozen in place, which is fine…if you’re the lion. Your fight or flight response is evolution’s way of telling you to move your ass if you don’t want to end up as kitty chow; it would be a pretty lousy survival mechanism if it did the equivalence of a PFC lobotomy every time you found yourself in harm’s way. The author could’ve used another expression or term to convey something more accurate—and less dramatic—like “downregulating” or “reducing the activity of” the PFC. Again…nuance, people.

It makes it nearly impossible to have productive conversations.

Learning to regulate your nervous system is a crucial skill for respectful dialogue.

Ok, be mindful of what you say when you disagree with someone if your intention is to have an open-ended, honest conversation with them. But what’s the point of this note? If the purpose was to reveal something worthwhile about the brain and the sympathetic system, I dare say the author missed the mark as the neuroscience it conveys is very weak. Otherwise, if the motive was to provide commentary about general etiquette and respect, then why have recourse to all that fluff about the brain? This is not neuroscience, ladies and gentlemen. This is garbage, which—if you’re a premium subscriber to the account in question—you pay to consume. I can’t imagine this kind of bunk being healthy for your brain…

Sometimes, these professional online grifters spread lies that can be downright dangerous. Take the following note, for instance, curtesy of Mr. As a neuroscientist:

For shame, Mr. As a neuroscientist, for shame!

I’m a neuroscientist.

You don’t say!

The best thing for your brain isn’t a pill.

Tell that to the person diagnosed with temporal lobe epilepsy who needs to take antiepileptic drugs to control their seizures, or the one suffering from Parkinson’s disease who depends on levodopa-based medications to bear with all the tremors and rigidity, or even someone who is at risk of stroke who must take anticoagulants to keep it at bay. This kind of one-size-fits-all, blanket statement is both unscientific—as it fails to distinguish between different population cohorts—and deleterious—as it can potentially cause actual harm if acted upon.

It’s morning sunlight. A long walk […]

Sure, buddy. Go take a hike, why don’t you?

REDEMPTION

In his final moments, Faust’s unyielding resolve vis-à-vis his altruistic vision of creating a free society for humanity is what ultimately redeemed his soul. After all that Mephistopheles had conjured up to lead him astray, Faust still actualized The Lord’s conviction that, even at his darkest, a good man cannot be deterred from the path of righteousness. And so, there is yet hope for the so-called neuroscientists whose notes I have used as case studies for the perversive pull of Substack’s Faustian bargain. To be fair, not everything that they—along with countless many other online grifters—have said or written is false or disingenuous, but a lot of it is, which is a testament to the intellectually and ethically corruptive sway that social media recognition can have on weak and fragile individuals who are morally flexible. Clout, money, fame, influence, power…these are all things that you can eventually acquire with the right combination of planning, execution, and luck, but once you bequeath your soul to the proverbial devil, you may never get it back. I’d be lying if I told you that the thought of dancing with said devil never crossed my mind—indeed it possesses undeniable magnetism for someone with my academic pedigree—but I didn’t go to the trouble of obtaining a PhD in cognitive neuroscience from one of the top institutions in the world just to peddle internet claptrap and become famous for it—no thanks! Instead, I remain steadfast in my commitment to sharing with you the science of the human brain as truthfully as I can within the limits of my knowledge, which, when it comes to personal integrity, is a lot more than I can say about pseudoscience peddlers, those with and without fancy University degrees and accolades. Like any self-respecting online writer or podcaster, I too am motivated by self-interest, but never to the point of trying to sell my subscribers, followers, and viewers half-baked concepts, redundant facts, and superfluous noise, if not brazen lies—which, nowadays, are stupidly easy to produce en masse thanks to cheap AI ghostwriters. So to those consuming this kind of bunk: don’t, you’re better off getting a lobotomy instead (I was going to say “eaten by a lion” but that’s too much). And to those promoting it: it’s not too late for a course correction; your soul—or at least what’s left of it—may depend on it.

Je ne me souviens de rien. Mes facultés mnémoniques m’ont désisté et j’ignore la raison. J’essaie ardemment de me concrétiser une idée, n’importe quoi, question de me réanimer l’esprit, mais je n’arrive pas à déloger le néant qui règne dans ma conscience. Même quand j’aboutis aléatoirement à une ébauche cognitive, au lieu de prendre forme en chose substantielle, celle-ci se transforme immédiatement en un poisson abstrait lisse qui me glisse entre les doigts des neurones.

Et pourtant, après plusieurs tentatives échouées, une notion parvient finalement d’éclore dans le creux de mes pensées : une image floue d’une âme qui s’échappe d’une bouche entrouverte. Serait-ce de cette manière que s’est garée ma mémoire, par l’évasion clandestine de mon esprit à travers l’espace entre mes lèvres? Suis-je en train de me faire des idées, de me façonner de fausses mémoires épisodiques? Réel ou pas, je me suis décidément désengourdit la cervelle. Quel soulagement!

Bon, il faut que je bouge un peu maintenant… Mais qu’est-ce qui me prend?

J’ai l’impression d’être paralysé; je ne peux guerre ni tourner la tête, ni soulever le bras. Tant qu’à mes jambes, je ne sais même pas si elles sont toujours là, attachées à ma hanche ou pas! À vrai dire, j’ai de la peine à percevoir mon entourage. Il y a une présence nébuleuse qui m’engloutit partiellement, ne me laissant qu’entrevoir un torse nu amaigri et des bras émaciés; pour tout ce qui est du nombril et des parties distales, mon champ visuel s’efface graduellement dans un infini obscur.

Mais qu’est-ce qui me prend? Où suis-je? Qui suis-je?

Je commence à avoir peur. Je transpire, la sueur me découle des tempes. Mon cœur élève de sa cadence. Ses battements deviennent de plus en plus forts, de plus en plus lourds et assourdissants. Quelle frayeur, quelle angoisse, quelle nausée! J’ai le cœur qui me monte jusque dans la gorge. Je vais le régurgiter.

Oui, j’ai la trouille. Je l’admets, je l’assume. Quel autre état d’âme suis-je supposé de me manifester en ce moment?

Je me suis réveillé - c’est clair - non pas d’un cauchemar, mais dans un cauchemar…

Pledge to ZAGROSCIENCE

A recent scientific review1 by my colleagues at the Montreal Neurological Institute and the Max Planck Institute for Human Cognitive and Brain Sciences offers a unifying perspective on these variations by focusing on system-level brain dynamics—particularly shifts in the brain’s sensory-to-transmodal information processing hierarchy, which is broken down in the video.

Join ZAGROSCIENCE

Read the source material here:

If you find the content here at ZAGROSCIENCE titillating, consider becoming a paid/founding member! 😉

Join ZAGROSCIENCE

Read the source material here:

“In this paper, we wanted to argue that 20 years of research is wrong!”

~ Dr. Mike X Cohen

In this episode of The Abstract, I speak with Dr. Mike X Cohen, the renegade of neuroscience! He is a former Associate Professor at the Donders Centre for Neuroscience in the Netherlands, where he conducted high-impact research in electrophysiology and systems neuroscience. Then, one day, he just quit! As he recounts in his article titled Why I left academia and neuroscience:

There is no single reason why I left. There were lots of reasons, and not any one of them was strong enough to have made me quit on its own.

Today, Dr. Cohen is on a mission to make complex mathematical and analytical topics accessible to everyone. He is the Founder of the popular educational platform, Sincxpress, where he creates online courses in applied mathematics, with a focus on data science, deep learning, statistics, linear algebra, calculus, signal processing, and scientific programming in MATLAB and Python. He is also a best-selling instructor at Udemy and the author of numerous textbooks. His self-professed objective is to equip individuals with the confidence to apply analytical skillsets effectively in practical, real-world scenarios.

In this podcast episode, Dr. Cohen and I break down a paper1 he and his colleagues published in 2020 in the Journal of Neuroscience titled Multiple Midfrontal Thetas Revealed by Source Separation of Simultaneous MEG and EEG. We also touch upon non-scientific themes like life outside academia, online charlatans, and his upcoming book about large language models (LLMs).

Where to find Dr. Mike X Cohen

Amazon books

Google Scholar

LinkedIn

Personal website

Sincxpress

Substack

Udemy

YouTube

Did you like this episode of The Abstract? Support my work by becoming a paid/founding member of my newsletter, ZAGROSCIENCE.

Pledge support

Read the source material here:

Back in July, I wrote an essay on the neural substrates of addiction in which I brought up Odysseus and his merciless 10-year-long voyage back to his home in Ithaca. I used his story as a metaphor to depict the drive and determination one needs to overcome substance use disorder, which, as I explained, consists of a vicious cycle of 3 phases that collectively hijack brain regions involved in executive function, motivation, reward, pleasure, and stress. As I was watching Christopher Nolan’s recent trailer of The Odyssey, it occurred to me that Homer’s eponymous epic, which is what the movie is based on, lends itself nicely to a discussion about a theme that is ubiquitous around this time of year: New Year’s resolutions. So, in this issue ZAGROSCIENCE, we bid 2025 farewell as we explore how you can optimize your resolve in 2026.

THE (NEURO)SCIENCE OF RESOLUTIONS

Overcoming the perils of the foreboding sea, the irresistible allure of Circe, and the unbreakable grip of the Cyclops—to say nothing of the furious anger of mighty Poseidon—requires the steadfast resolution of an unyielding mind at the helm of a stalwart body. And yet, not even the mighty Odysseus could’ve accomplished all his feats alone, which speaks to the necessity of adequate support for staying the course.

In 2020, Oscarsson and colleagues published one of the largest studies on New Year’s resolutions, investigating a sample of 1,066 participants randomized into 3 groups: active control (AC), some support (SS), and extended support (ES). The objective was to categorize different kinds of resolutions, evaluate the success rate of each, and establish the importance of support in adhering to one’s plans. The AC group received brief guidelines and 3 follow-ups throughout the year; the SS group was given social support instructions, 12 monthly follow-up assessments, and one email containing exercises and coping strategies; the ES group was provided specific goal instructions, interim objectives, 12 monthly follow-ups, and 4 emails with exercises. Overall, the top 3 resolutions that participants chose pertained to physical health (33%), weight loss (20%), and eating habits (13%), with a general success rate of 55% across all groups over the course of 1 year.

One major finding in this study was that approach-oriented goals showed higher rates of adherence than avoidance-oriented goals (58.9% vs. 47.1%). An approach-oriented objective is one in which you are actively trying to move towards a positive outcome, whereas in an avoidance-oriented paradigm, you are attempting to steer clear of a negative consequence. Approach-oriented strategies are neurally associated with networks linked to reward, positive emotions, and flexible control, whereas avoidance-oriented goals preferentially recruit brain systems involved with anxiety, vigilance, and cognitive control.

In a recent neuroimaging study (Sakaki et al., 2024), researchers examined how different objective-driven strategies for the same task can modulate the brain’s reward system, namely, the striatum and ventromedial prefrontal cortex (vmPFC). A total 20 young adults performed a stopwatch task while being scanned in the MRI machine. At each trial, participants had to pause a stopwatch via button press within 50 ms of the 5-second time point. Depending on whether or not the stopwatch was paused within the allotted time window, the trial was either deemed a success or a failure. To assess the effect of the subject’s motivational state of mind, participants completed the task in 3 separate blocks: in the approach block, they gained points for succeeding and lost nothing for failing; in the avoidance block, they lost points for failing and gained nothing for succeeding; in the control block, they simply watched the timer and pressed a button without gaining or losing points.

Univariate analysis was used to compare the different conditions while functional connectivity determined network associations during different goal states. Additionally, to compare the similarity of brain patterns across the different conditions, a multivariate technique known as representational similarity analysis (RSA) was performed. Behaviourally, participants reported relatively higher levels of excitement and engagement in the approach condition while exhibiting more anxiety and disappointment during avoidance. Intriguingly, regardless of emotional differences in the two conditions, the brain’s reward network showed comparable activity. Even so, neural signatures in the striatum and vmPFC were found to differentiate between approach and avoidance strategies, speaking to a fundamental distinction in how this neural circuitry processes them. What’s more, regions outside canonical reward areas showed different activation effects of success/failure between experimental conditions, with significant interactions in the lateral occipital gyrus, hippocampus, cerebellum, and brainstem. These brainwide variations may be an important piece of the puzzle in understanding why approach-oriented goals tend to feel more enjoyable, sustainable, and effective in driving behaviour than avoidance.

Another interesting result by Oscarsson et al., (2020) is the nonlinear relationship between support and resolution success rate. The report found that the SS cohort, which had received moderate support, saw significantly greater success in adherence to New Year’s resolutions than either the AC group—which had received almost no aid—or the ES group—which was given extensive and prolonged help. The authors suggested that too many goals and interim deadlines may actually increase the potential for failure. Indeed, clearly formulated time-sensitive objectives are a great way to recruit and mobilize your energy and cognitive resources, but too many of them within a relatively short period of time could end up backfiring by overburdening you with noise and redundancy.

The article also noted some encouraging secondary outcomes. Namely, participants who adhered to their resolutions throughout the year scored significantly higher in assessments of self-efficacy and subjective quality of life than those who gave up. Furthermore—and perhaps more importantly—regardless of whether or not their New Year’s resolutions were successful, participants generally showed a positive change in reducing procrastination, which speaks to the potential psychologically beneficial secondary effects of harnessing one’s motivation and drive irrespective of ultimate outcome.

TAKEAWAYS

1. Set realistic goals in which you are ideally moving towards a specific positive outcome. This approach will keep you motivated on the path ahead while possibly strengthening brain networks involved in feeling accomplished.

2. Keep track of your progress but don’t micromanage everything to the point where you overburden yourself with increased failure potential. There is a fine line here, which only you can identify for yourself.

3. If you falter, it’s ok—you’re only human. But try and resolve the root causes of your shortcomings. Triumphing over temporary setbacks along the way can actually make you more resilient against inertia post-failure, which is good!

4. Just like Odysseus, you need some support—not too much, but not too little either…perhaps a close confident who checks up on your progress once in a while or even a self-made memo that reminds you of the larger picture when you’re feeling down.

So whatever you set your sights on for 2026, whether it’s becoming more fit, eating healthier, or making millions of dollars with your newsletter, don’t forget to make your resolution simple and clear so that it counts. And remember, you don’t need to become a beast of burden to put your ass into it!

Let me know if you value all the work I’ve done this year at ZAGROSCIENCE

Pledge support

Some relevant essays from the ZAGROSCIENCE vault.

If you’re into neuroscience (the real deal - not the cheap online fluff), check out my new podcast The Abstract, where I interview the women and men who are driving it forward.

REFERENCES

Oscarsson, M., Carlbring, P., Andersson, G., & Rozental, A. (2020). A large-scale experiment on New Year’s resolutions: Approach-oriented goals are more successful than avoidance-oriented goals. PLoS One, 15(12), e0234097.

Sakaki, M., Murayama, K., Izuma, K., Aoki, R., Yomogita, Y., Sugiura, A., Singhi, N., Matsumoto, M., & Matsumoto, K. (2024). Motivated with joy or anxiety: Does approach-avoidance goal framing elicit differential reward-network activation in the brain?. Cognitive, affective & behavioral neuroscience, 24(3), 469–490. https://doi.org/10.3758/s13415-024-01154-3

“Hopefully I can help other families that are going through the same thing my family and I went through.”

~ Dr. Sara Larivière

In this episode, I speak with Dr. Sara Larivière, Assistant Professor at the Department of Nuclear Medicine and Radiobiology at Université de Sherbrooke. She is at the head of the SLIC (Sherbrooke Laboratory for Integrative Connectomics), holding a Canada Research Chair in Neuroimaging of Epilepsy. Dr. Larivière and I are friends and colleagues, having received our doctorate training at the same lab under the supervision of Dr. Boris Bernhardt, my first guest on the show (see the link below).

Her lab integrates neuroimaging, computational statistics, machine learning, and pipeline engineering to advance knowledge of the human brain. By developing and deploying neuroinformatics and data science tools, her team investigates normal and atypical brain development, with the goal of enhancing personalized prognosis and therapies, ultimately aiming to shape public health policy.

In this instalment, Dr. Larivière and I break down a 2024 article1 we published in the Journal of the American Medical Association - Neurology, titled Brain Networks for Cortical Atrophy and Responsive Neurostimulation in Temporal Lobe Epilepsy. We discuss how accounting for widespread cortical dynamics in temporal lobe epilepsy (TLE) may inform novel therapeutic approaches that move beyond traditional subcortical targets.

INTERNAL LINKS

Epilepsy:

Connectomics:

Neurodivergence/autism:

EXTERNAL LINKS

Email: sara.lariviere@usherbrooke.ca

SLIC homepage: https://slic-lab.github.io//

Université de Sherbrooke profile: https://www.usherbrooke.ca/recherche/fr/specialistes/details/sara.lariviere

Google Scholar profile: https://scholar.google.com/citations?user=Fet1xBYAAAAJ&hl=en

KEEP THOSE NEURONS FIRING

Did you like this episode of The Abstract? Support my work by becoming a paid/founding member of ZAGROSCIENCE.

Pledge support

“Thou shalt not scathe thine neurones.”

~ The Unspoken Commandment

IRON SHARPENS IRON

I come from a strict mixed martial arts pedigree: gruelling sessions of conditioning, relentless drilling of technique, and hard bouts of grappling and sparring. When Sifu1 asks you to jump, your only thought is, “how high?” There is no room for physical fatigue or mental weakness—it’s full commitment to the here and now in body and mind, or nothing else. Mine is a tried and tested warrior’s paradigm in which iron sharpens iron. And yet, while 18-year-old me thrived in this hardened environment, 40-year-old me today can certainly impart a few words of wisdom on my younger self. So if you’re anything like 18-year-old me—indefatigable and single-minded but sometimes lacking perspective—strap yourself in, because in this issue of ZAGROSCIENCE, class is in session and the master is taking you to school!

NEUROBIOLOGICAL OVERVIEW

A recent review published in Brain Research by Gazerani (2025) frames skill learning as an iterative adjustment in cortical and subcortical circuits that minimizes prediction errors while maximizing reward. This process is similar to how machine learning algorithms reduce the cost (or loss) function, which includes making predictions, weighing these predictions against defined targets, and recalibrating system parameters to ensure that downstream predictions are closer to expected values. Inside the brain, this mechanism is sustained by a convergence of processes involving neuroplasticity (e.g., structural and functional reorganization), time-dependent consolidation, optimal cognitive load, and neuromodulation via signalling compounds like dopamine and norepinephrine.

PROGRESSIVE COGNITIVE LOAD FOR OPTIMIZED SKILL ACQUISITION

In Christiansen et al., (2020), researchers compared two groups on a visuomotor task over the course of 6 weeks. In the progressive training (PT) cohort, training difficulty was continuously readjusted to match individual skill levels whereas it was held constant in the nonprogressive training (NPT) group. The study found that, on average, participants adhering to the PT regimen outperformed their NPT counterparts by a factor of 2, with associated heightened corticospinal excitability. Two additional retention reassessments were conducted after the original 6-week period. At retention 1, following 8 days of discontinued practice, group differences were replicated, showing that PT had maintained its performance advantage over NPT. However, when both cohorts were reappraised again after 14 months of no training, scores no longer differed, speaking to the importance of sustained practice in honing a skill. Nonetheless, the authors remarked that when it comes to learning, a regimen that offers gradually more challenging tasks dynamically tailored to the individual’s proficiency may be superior to one in which the challenge remains constant:

Our study demonstrates for the first time that an individually adjusted, progressive training protocol involving 6 weeks of visuomotor skill acquisition results in significantly better motor performance compared to a non-progressive skill learning protocol. Progressive training was also accompanied by pronounced increases in corticospinal excitability suggesting that continuous challenge during motor training drives corticospinal plasticity.

SPACING OUT SESSIONS FOR CONSOLIDATION

Time-dependent consolidation is when fleeting neural traces become integrated into longer-lasting memory circuits. A new study by Yang et al., (2025) found that spaced learning may be crucial for the formation of durable memories. Unlike mass learning, which is marked by cramming of information—over a relatively short period of time—spaced learning involves the distribution of study sessions over a longer timeframe. The researchers used a between-subject, day-based design where 48 participants were randomly allocated to either a 3-day spaced learning cohort or a 1-day massed learning group as they learned 60 picture-word pairs. Resting-state and task-based fMRI data were acquired during multiple delayed tests: immediately, after 1 week, and after 1 month. To examine neural integration and spontaneous memory replay, the study employed representational similarity analysis (RSA) in the hippocampus and regions of the default mode network (DMN), with complementary multivariate analytical techniques (e.g., multivoxel functional connectivity and multivariate pattern analysis). Memory performance was found to be significantly higher in the spaced learning cohort relative to the massed learning group in the 1-week and 1-month delayed tests, but not in the immediate test. Spaced learning was further associated with superior retention rates across both delays. Neural integration in the dorsal-medial and medial-temporal DMN at 1-month was linked to durable memory induced by spaced learning. Increased neural replay was additionally observed in the dorsal-medial DMN. Overall, the authors concluded that while the hippocampus is heavily involved in initial memory processing, time-dependent consolidation may occur as a result of enhanced integration in the neocortex rather than the hippocampus.

SLEEP FOR BRAIN RECALIBRATION

In 2021, Newbury et al. published a dual meta-analysis that examined the negative effects of acute sleep deprivation on memory both before and after learning (e.g., BL vs. AL), spanning 50 years of research (e.g., from 1970 to 2020). In the BL module, a total of 31 studies were quantitatively analyzed, yielding a medium-to-large detrimental effect size (Hedges’ g = 0.621, 95% CI [0.473, 0.769]), supporting the idea that sleep prepares the brain for new learning. In the AL module, 45 studies were examined, resulting in a smaller, albeit non-negligible, effect size (Hedges’ g = 0.277, 95% CI [0.177, 0.377]), indicating that sleep plays a role in the consolidation of newly learned information. Interestingly, the negative effects of sleep deprivation were more significant for procedural memory tasks (e.g., finger-tapping, visual discrimination, mirror tracing, etc.) compared to declarative tasks (e.g., word pairs, route learning, categorization, etc.). What’s more, this difference was more noticeable if testing occurred immediately after deprivation rather than after recovery sleep, pointing to a restorative effect of sleep following deprivation. Generally speaking, procedural memory taps into motor and/or visual processing, whereas declarative memory is primarily hippocampal-dependent. Procedural memories are believed to require immediate offline (e.g., sleep) consolidation as they are unlikely to rely on online (e.g., awake) hippocampal-neocortical dynamics. As for declarative memories, while they are incidentally encoded by the hippocampus during wakefulness, slow-wave sleep (SWS) is believed to support their strengthening and integration into the neocortex via repeated reactivation. In both cases, sleep ultimately plays an important factor.

MULTIMODAL LEARNING FOR MAXIMAL INTEGRATION

Multimodal enrichment is when complementary sensory and/or motor stimuli are incorporated into the learning process. Imagine you’re learning to taste and identify wines. In a unimodal paradigm, an instructor may simply have you taste the wine and then verbally describe its aroma (e.g., “This wine presents with an elegant bouquet of violets.“). A multimodal approach may, in addition to the verbal description (e.g., auditory stimulus), have you taste the wine (e.g., gustatory stimulus) and also smell a violet flower (e.g., olfactory stimulus) for cross-comparison. A review by Mathias & von Kriegstein (2023) integrated cognitive, neuroscientific, and computational approaches to shed light on the effectiveness of multimodal enrichment on skill acquisition, noting that it can drastically improve learning across multiple domains (e.g., vocabulary acquisition, reading, mathematics, etc.). The review found that behavioural benefits were associated with underlying cortical changes: sensory and motor areas exhibited crossmodal integration in response to enriched items during subsequent unimodal tasks. Findings supported the idea that multimodal enrichment contributes to enhanced multisensorial consolidation at the level of the brain, significantly reinforcing cortical representations that support learning and memory.

TAKE-HOME MESSAGE

Based on the latest in neuroscientific research, optimization of skill acquisition can be broken down into the 4 points:

1. individually-tailored progressive cognitive load to operate at optimal capacity (e.g., “stretch zone”)

2. spaced out sessions to enhance consolidation of new skills in between bouts

3. good quality sleep to promote skill/knowledge integration

4. multimodal enrichment learning to activate converging neural pathways

Implementing this multipronged approach when learning new abilities can maximize how quickly and reliably you can expand your arsenal of procedural and/or declarative expertise.

THE EDICT OF LONGEVITY

I remember Sifu telling me how he was in the best shape of his life in his mid-40s. As someone who is approaching this age range myself, I find the prospect of having my best years still ahead of me reinvigorating. Of course, I need to be proactive about it. Although I’m no longer training martial arts anywhere near the frequency and intensity I would like to be, I’ve taken up weightlifting in the last couple of years as a substitute. I have implicitly adhered to a semi-structured regimen that incorporates, where applicable, the 4 points mentioned earlier—although good sleep is not always easy to come by when you have a toddler at home… Nonetheless, while my cardiovascular fitness remains a work in progress, I have never been physically stronger than what I am now (see the video below where I deadlift 315 lbs [~143 kg] for 6 repetitions with relative ease).

In retrospect, the biggest piece of advice I’d give my younger self would be to ease off the full contact sparring. It goes without saying that minimizing concussive blows to the head is a good strategy for safeguarding the integrity of the primary organ responsible for learning. If your brain is compromised in the process of acquiring and honing a skillset, the process itself becomes compromised. In fact, I’m not convinced that full contact sparring is a necessary training paradigm even for pugilistic professionals like mixed martial artists and boxers. As discussed earlier, your brain thrives under optimal conditions of cognitive load: too little input, and it’s operating at subpar levels; too much, and it’s overburdened—and getting knocked unconscious is your brain’s way of spelling out “system overload!” Indeed, repetitive hard blows to the cranium serve no other purpose than to increase your likelihood of developing chronic traumatic encephalopathy (CTE). Imagine smashing your computer with a baseball bat in an attempt to update its software—it just makes no sense! But based on what neuroscience and martial arts have taught me, it’s an analogy that rings true with giving yourself brain damage while aspiring to sharpen your combat skills. In my opinion, more rounds of light sparring, technique drilling, reaction training, and mental imagery constitute a far more auspicious regimen than harming your noggin. So the next time you try your hands at something new, especially something that requires physical exertion, keep the following unspoken commandment in mind: “Thou shalt not scathe thine neurones.” Abiding by this simple and self-evident edict will enable you to thrive for years to come while continuing to learn and grow as an individual.

REFERENCES

Christiansen, L., Larsen, M. N., Madsen, M. J., Grey, M. J., Nielsen, J. B., & Lundbye-Jensen, J. (2020). Long-term motor skill training with individually adjusted progressive difficulty enhances learning and promotes corticospinal plasticity. Scientific reports, 10(1), 15588.

Gazerani, P. (2025). The neuroplastic brain: current breakthroughs and emerging frontiers. Brain Research, 149643.

Goldberg, H. (2022). Growing brains, nurturing minds—neuroscience as an educational tool to support students’ development as life-long learners. Brain sciences, 12(12), 1622.

Mathias, B., & von Kriegstein, K. (2023). Enriched learning: Behavior, brain, and computation. Trends in Cognitive Sciences, 27(1), 81-97.

Newbury, C. R., Crowley, R., Rastle, K., & Tamminen, J. (2021). Sleep deprivation and memory: Meta-analytic reviews of studies on sleep deprivation before and after learning. Psychological bulletin, 147(11), 1215.

Yang, Y., Huang, Z., Yang, Y., Fan, M., & Yin, D. (2025). Time-dependent consolidation mechanisms of durable memory in spaced learning. Communications Biology, 8(1), 535.

Subscribe now

Watch the 1^st episode of The Abstract right now, where I speak with world-renowned neuroscientist, Dr. Boris Bernhardt, about human brain function and its relation to local and brainwide cortical dynamics.

“Don’t let perfect get in the way of good enough.”

~ Dr Boris Bernhardt

In this inaugural episode, I speak with one of the world’s foremost experts in neuroimaging and brain connectomics, Dr Boris Bernhard, Professor of Neurology and Neurosurgery at McGill University and Canada Research Chair in Cognitive Neuroinformatics. Dr Bernhardt is the Principal Investigator at the Multimodal Imaging and Connectome Analysis (MICA) Lab situated at the McConnell Brain Imaging Centre of the Montreal Neurological Institute (aka, The Neuro). He is also my friend and professional colleague, having served as my thesis supervisor during my PhD studies at McGill University.

His lab investigates how the organization of structural and functional brain networks supports integrated cognition in both healthy individuals and those with disorders, particularly epilepsy and autism. His team has played a key role in recent conceptual and methodological breakthroughs for mapping spatial patterns in the brain as continuous gradients—especially hierarchical ones that span from sensory-motor regions to higher-order transmodal systems responsible for associative processing. His research routinely combines multimodal neuroimaging, network neuroscience, 3D histology, and transcriptomics.

On this occasion, Dr Bernhardt and I discuss a recently published paper1 by his lab in Nature Communications, titled Multimodal gradients unify local and global cortical organization. We break down how brain function is situated at the nexus of local and global cortical dynamics.

EXTERNAL LINKS

MICA homepage: https://mica-mni.github.io/

The Neuro profile: https://www.mcgill.ca/neuro/boris-bernhardt-phd

Google Scholar profile: https://scholar.google.ca/citations?user=WlKrpE4AAAAJ&hl=en

Themes covered include:

• prevalence & economic burden

• treatment options

• disruptions to the functional web of the brain

• Dante’s account of the Inferno

Read the full source material here:

Subscribe now

“[T]he real mystery of absolute pitch is not why some people possess this ability, but instead why it is so rare.”

~ Diana Deutsch

Should I read this article?

Neurodivergence is a catchall term that describes atypical mental predispositions. In the scientific literature, it is oftentimes used in the context of individuals with autism—or autistic spectrum disorder (ASD). Associated with altered cognitive, behavioural, and social dynamics, ASD is a neurodevelopmental condition featuring heterogenous clinical findings. Recently, my colleagues at the Montreal Neurological Institute and the Max Planck Institute for Human Cognitive and Brain Sciences published a scientific review that addressed these disparities through the unifying lens of system-level brain dynamics, which I will examine in depth. If you’re neurodivergent yourself, know someone who is, or have ever wondered about what neurodivergence entails from an information processing perspective, you’ll definitely enjoy this read.

Should I read your neuroscience articles in general?

I obtained my PhD in cognitive neuroscience from McGill University in 2024, specializing in memory and consciousness. I have extensive research experience at the Montreal Neurological Institute (The Neuro) where I conducted neuroimaging studies that tapped into human cognition in healthy individuals and patients with epilepsy. I have also (co-)authored over 25 peer-reviewed publications in leading scientific journals like Nature Neuroscience, Cerebral Cortex, Brain, and Neuroimage. Beyond bringing a strong academic foundation to my work here on Substack, I always aim to write clearly and effectively to keep you informed about and engaged with brain science. You may also find themes drawn from philosophy, literature, history, or mythology woven into my writing for added depth.

Should I subscribe to ZAGROSCIENCE (or upgrade my current membership)?

If you’re interested in hardcore neuroscience with an artistic flair, ZAGROSCIENCE will challenge your mind and appeal to your senses. It includes an expanding collection of audiovisual and written/narrated essays exploring memory, consciousness, psychedelics, epilepsy, addiction, and much more. I’m also launching The Abstract, a new podcast featuring interviews with leading brain researchers, including some of my professional colleagues. Only valued ZAGROSCIENCE patrons can submit questions to be shared with guests on the show.

Subscribe now

ABSOLUTE PERFECTION

In The Enigma of Absolute Pitch (2006), Diana Deutsch examines the emergence of the namesake phenomenon, also known as perfect pitch, the capacity to correctly name any musical note when it is played alone. With enough practice especially at a young age, almost anyone can acquire the ability to identify a note when provided with a referential sound for comparison, but even professional musicians who do not possess absolute pitch will fail to do so in the absence of an established baseline. Deutsch gives an example:

…most musicians, when presented with the note F and given its name, have no difficulty in naming the note two semitones higher as G, the note four semitones tones higher as A; and so on.

The author remarks that unlike musical notes, one does not require a reference colour to correctly name another one. Barring underlying issues like colour anomia or blindness, we can promptly recognize isolated hues with no effort. Deutsch, who claims to possess absolute pitch herself, suggests that the capacity to accurately name isolated acoustic tones is, in a practical sense, similar to identifying individual colours separately. With a prevalence of less than one in ten thousand in the general population, this comparison between notes and colours may leave you wondering why perfect pitch is so rare. Equally compelling is how there is a unique neurodivergent demographic in which these intriguing perceptual phenotypes (e.g., absolute pitch and synesthesia) are disproportionately more prevalent compared to neurotypical cohorts. In this issue of ZAGROSCIENCE, we will explore the network-level idiosyncrasies that make the neurodivergent brain, especially in regard to autism, such a distinctive biological hardware when it comes to information processing.

SOME PERSPECTIVE

Grokipedia defines autism as follows:

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by persistent deficits in social communication and interaction across multiple contexts, as manifested by difficulties in social-emotional reciprocity, nonverbal communicative behaviors, and developing, maintaining, and understanding relationships, alongside restricted, repetitive patterns of behavior, interests, or activities, and sensory hyper- or hyporeactivity.

The 2019 Canadian Health Survey on Children and Youth (CHSCY), a nationwide census that collected information pertaining to the health and well-being of the country’s non-adult population, found that 2% of respondents (e.g., 819 out of 39,951) presented with ASD. More than half of all diagnoses (e.g., 53.7%) had occurred before the age of five, with males comprising approximately 80% of total cases. The survey also noted that 68.7% had long-term comorbidities—the most common ones being “attention deficit disorder/attention deficit hyperactivity disorder, learning disability/disorder and anxiety disorder […]” Moreover, challenges in communication, adapting to change, socializing with friends, and other daily struggles were reported in 73.3% of those between the ages of 2 and 17. The Centers for Disease Control and Prevention (CDC) reports similar statistics in the United States, with 3.2% of children aged 8 years identified as having ASD and a threefold prevalence among boys relative to girls. ASD diagnosis can be complicated as there are no specific medical tests for its detection. Instead, approaches involve ongoing developmental monitoring and screening. Moreover, there is no “treatment” per se, but only interventions that seek to lessen the negative impact of symptoms on daily functioning and quality of life.

REVOLVING ON THE SENSORIMOTOR-TRANSMODAL AXIS

Notwithstanding the tremendous scientific efforts under way, the neural mechanisms underlying cognitive, behavioural, and social changes in ASD remain unsettled, which can be largely attributed to its heterogenous phenotypical manifestation. Descriptive models have tackled the issue from different angles, with impaired Theory of Mind and altered perception among proposed explanations for certain components of ASD symptomatology, but none fully grasping its multidimensional complexity. It has been suggested that a synthesis of underlying cortical dynamics can be achieved through careful consideration of a key feature of the brain: its hierarchical organization with respect to information processing—which is the main theme of a newly published review by Bernhardt1 et al., (2025). In this work, the authors describe ASD in the context of large-scale network dynamics, defined by an atypical computational neural architecture.

Network neuroscience has shown that information processing in the human brain is organized into a hierarchical architecture consisting of separate axes known as gradients (Margulies et al., 2016). Each brain gradient is defined by smooth spatial transitions along a spectrum anchored at each end by a macroscale network. Gradients are non-overlapping components derived from functional connectivity data when applying nonlinear dimensionality reduction algorithms (e.g., diffusion embedding). The first gradient, which recapitulates the largest share of total information in the brain’s functional connectome, differentiates unimodal regions of the sensory/motor system from transmodal areas of the default mode network (DMN), hence its designation of sensorimotor-transmodal (S-T) axis. The next largest variance in the data is explained by the second gradient, which captures variations within the unimodal end of the S-T axis, with the visual cortex on one extreme and the somatomotor/auditory cortex on the other. Importantly, these principal gradients have been validated using other modalities, including structural neuroimaging, histology, and gene expression.

If you want to learn more about brain gradients, read/listen to my following article:

Based on recent neuroimaging and connectomics analyses (Hong et al., 2019), the review by Bernhardt et al. noted a relative reduction in functional differentiation along the S-T axis in ASD. This observation can be visualized in the gradient spaces below where, compared to neurotypical controls (top left), individuals with autism (top right) display a “squeezed” S-T component, illustrating their neurodivergence along the most informative functional connectivity gradient in the human brain.

Speaking to these atypical S-T dynamics, Bernhardt et al. also pointed to a lack of convergence of information flow from unimodal to transmodal areas in ASD. Using stepwise connectivity analysis, which iteratively tracks how a functional connectome evolves from previously defined brain “seeds” to other regions, Hong et al., (2019) showed that cortical maps derived for visual (V1), auditory (A1), and somatosensory (S1) areas failed to reach transmodal DMN regions in autistic individuals even after 200 steps. As can be seen in the above figure, arrows arising from V1, A1, and S1 seeds converge near the DMN peak in controls (bottom left), which is clearly not the case in the autistic group (bottom right).

A system-level understanding of ASD-associated behavioural/cognitive alterations can additionally help consolidate insights from different predictive models of the condition. Bernhardt et al. mentions that theoretical frameworks rooted in conditional probabilities have suggested a reduction in top-down regulation in the processing of perceptual information. Generally speaking, our brains are constantly fine-tuning previously consolidated models of how the world is. These so-called cognitive priors may be affected in ASD. The authors write:

According to Bayesian models, the influence of prior experiences on perception is reduced in autism. This is possibly due to either a devaluation of prior knowledge or an enhanced weighting of sensory input and leads to perceptions of the world that are atypically grounded in realism […]

The review also notes alternative mechanisms, such as one-size-fits-all priors:

On the other hand, it has been argued that the predictive framework may artificially isolate a particular narrative and then generalize it to a range of perceptual manifestations, some of which are unimpaired. For example, there is evidence that individuals with autism perform comparably to neurotypical individuals when learning statistical regularities in a task that requires predictive learning […]

If you want to learn more about cognitive priors and how they shape our conscious perceptions, read my following essay:

Impairment in social functioning is another area that systems neuroscience may elucidate. The Autism Diagnostic Observation Schedule (ADOS), for instance, is a standardized assessment protocol that is widely used in ASD diagnosis. It comprises various activities, such as play or conversation, that help assess social skills, communication, and behaviour. Bernhardt et al. relates the following:

Indeed, cortex-wide functional gradients and stepwise connectivity patterns were found to predict total [ADOS] scores, as well as subscores for social interactions and repetitive and restricted behaviors, even when cross-validation was used […] Predictive features were observed in the default mode network, as well as sensory systems, indicating that alterations in both sensory and higher-order regions of the cortical hierarchy may synergistically contribute to autism-related behaviors.

While the review acknowledges several outstanding questions, such as the extent to which system-level alterations can distinguish between different forms of neurodivergence, it establishes neural hierarchies involved in information processing as a unifying framework for understanding the multidimensional complexities that manifest in ASD.

CONNECTING THE DOTS

ASD is a heterogeneous neurodevelopmental condition affecting about 2-3% of children, predominantly males, with early-onset diagnoses, high comorbidity rates (e.g., ADHD, anxiety), and no definitive medical tests or cures—only interventions to mitigate daily challenges in communication, socialization, and adaptation. A 2025 review by Bernhardt et al. frames ASD through network neuroscience, highlighting atypical dynamics along the brain’s primary S-T functional gradient, where autistic individuals show reduced differentiation and failed convergence of information flow from unimodal sensory areas (e.g., visual, auditory, somatosensory) to transmodal DMN regions, as evidenced by compressed gradient spaces and stepwise connectivity analyses. This systems-level disruption aligns with Bayesian models suggesting diminished top-down influence over perception or one-size-fits all cognitive priors, potentially explaining hyper-realistic subjective experiences, while gradient patterns predict ADOS diagnostic scores for social and repetitive behaviours, unifying sensory and higher-order cortical alterations into a hierarchical framework for the multidimensional symptomatology of ASD.

STROKE OF GENIUS

The opening quote in this article posed the question of why perfect pitch is so rare. Considering just how readily one can name isolated colours, it does seem a bit odd that less than 0.01% of the general population can do the same for musical notes. And yet, its prevalence among neurodivergent individuals is far superior than this baseline. As per Bernhardt et al.:

Synesthesia and absolute pitch […] are about 100 times more prevalent in autistic individuals than in neurotypical individuals […]

This oddity is further highlighted by the possibility, although never confirmed, that the two most well-known musical geniuses of the 18^th and 19^th centuries, Mozart and Beethoven—who possessed perfect pitch—could have been autistic. The first account of ASD came much later, describing the curious case of a patient by the name of Donal Triplett (Kanner, 1943). Intriguingly, it is highly likely that Triplett also possessed absolute pitch. According to the original article by Kanner:

At the age of 1 year “he could hum and sing many tunes accurately.”

An article by The Atlantic (Donvan & Zucker, 2010) describes Triplett’s musical prowess as follows:

Oddly distant, uninterested in conversation, and awkward in his movements, Donald nevertheless possessed a few advanced faculties of his own, including a flawless ability to name musical notes as they were played on a piano […]

As with almost all things relating to human nature, the truth about the general scarcity of absolute pitch and its higher prevalence in specific cohorts is to be found at the intersection of evolution, genetics, biology, and environment. A system-level account of the autistic brain, while still in its infancy, holds great promise for advancing our understanding of neurodivergence, from its many burdens to its multiple blessings.

I’m neither neurodivergent nor possess perfect pitch, but I do have a musical ear and obliging fingers!

Share

Subscribe now

Follow me on X

Leave a comment

Add me on LinkedIn

Share ZAGROSCIENCE

Check out ZAGROSCIENCE on YouTube

REFERENCES

Bernhardt, B. C., Valk, S. L., Hong, S. J., Soulières, I., & Mottron, L. (2025). Autism-related shifts in the brain’s information processing hierarchy. Trends in Cognitive Sciences.

Canada, P. H. A. of. (2022, February 3). Autism spectrum disorder: Highlights from the 2019 Canadian health survey on children and youth. Www.canada.ca. https://www.canada.ca/en/public-health/services/publications/diseases-conditions/autism-spectrum-disorder-canadian-health-survey-children-youth-2019.html#a1

CDC. (2024, February 22). Autism Spectrum Disorder (ASD). Autism Spectrum Disorder (ASD). https://www.cdc.gov/autism/

Deutsch, D. (2006). The enigma of absolute pitch. Acoustics Today, 2(4), 11.

Donvan, J., & Zucker, C. (2010, August 30). None. The Atlantic; The Atlantic. https://www.theatlantic.com/magazine/archive/2010/10/autisms-first-child/308227/

Government of Canada, S. C. (2023, February 16). Canadian Health Survey on Children and Youth (CHSCY). Www23.Statcan.gc.ca. https://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey&SDDS=5233

Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2, 217–250.

Margulies, D. S., Ghosh, S. S., Goulas, A., Falkiewicz, M., Huntenburg, J. M., Langs, G., ... & Smallwood, J. (2016). Situating the default-mode network along a principal gradient of macroscale cortical organization. Proceedings of the National Academy of Sciences, 113(44), 12574-12579.

Hong, S. J., Vos de Wael, R., Bethlehem, R. A., Lariviere, S., Paquola, C., Valk, S. L., ... & Bernhardt, B. C. (2019). Atypical functional connectome hierarchy in autism. Nature communications, 10(1), 1022.

“My mind, because of its disdainful temper, believing it could flee disdain through death, made me unjust against my own just self.”

~ Dante Alighieri (Divine Comedy - Inferno)

Should I read this article?

Depression is a global crisis, impacting over 300 million people and contributing to countless suicides, while also imposing a massive economic burden beyond healthcare costs. Understanding the neural underpinnings of depression is a critical goal in medicine, as it can improve early detection, diagnosis, and intervention. Functional network analysis, a key tool in modern neuroscience, offers great promise for uncovering vital insights into the nature of depression and strategies for its prevention or treatment. Whether you’ve faced depression yourself, know someone who has, or are simply curious about the topic, this article should speak to you. It provides a crucial perspective for understanding what this complex condition is by exploring it through the lens of network neuroscience.

Should I read your neuroscience articles in general?

I obtained my PhD in cognitive neuroscience from McGill University in 2024, specializing in memory and consciousness. I have extensive research experience at the Montreal Neurological Institute (The Neuro) where I conducted neuroimaging studies that tapped into human cognition in healthy individuals and patients with epilepsy. I have also (co-)authored over 25 peer-reviewed publications in leading scientific journals like Nature Neuroscience, Cerebral Cortex, Brain, and Neuroimage. Beyond bringing a strong academic foundation to my Substack articles, I aim to write clearly and effectively to keep you informed about and engaged with brain science. You may also find themes drawn from philosophy, literature, history, or mythology woven into my writing for added depth and context.

Should I subscribe to ZAGROSCIENCE (or upgrade my current membership)?

If you’re interested in hardcore neuroscience with an artistic flair, ZAGROSCIENCE will challenge your mind and appeal to your senses. It includes an expanding collection of audiovisual and written/narrated essays exploring memory, consciousness, psychedelics, epilepsy, addiction, and much more. I’m also launching The Abstract, a new podcast featuring interviews with leading brain research experts, including some of my professional colleagues. Only valued ZAGROSCIENCE patrons can submit questions to be shared with guests on the show.

Subscribe now

THE PUZZLED CORTEX

My soon-to-be three-year-old daughter displays remarkable visuospatial awareness when absorbed in one of her favourite activities: doing jigsaw puzzles with daddy! This task requires cognitive flexibility, integrating multiple low-level features (e.g., pattern, shape, and size) across puzzle pieces for correct assembly. Occasionally, she tries to fit two incongruous pieces together before finding the compatible one for each. Watching her resolve such complex problems sometimes reminds me of how the human brain’s functional networks must work harmoniously for optimal performance. What happens to a jigsaw set when jamming mismatching puzzle pieces together is oddly reminiscent of disruptions in this dynamic web of neurons, such as those caused by clinical depression. In both cases, the result is marked by incoherence and instability—and in this issue of ZAGROSCIENCE, we will explore the network dynamics involved in the latter.

DEFINING THE PROBLEM

According to Chand & Arif, (2025), depression can be defined as “a mood disorder that causes a persistent feeling of sadness and loss of interest,” which can be divided into 5 categories: disruptive mood dysregulation disorder, major depressive disorder (MDD), persistent depressive disorder (dysthymia), premenstrual dysphoric disorder, and depressive disorder due to another medical condition. Core features of depression include sadness, emptiness, or irritability, along with somatic and cognitive maladjustments that substantially disrupt an individual’s ability to function. The WHO estimates that 5.7% of adults around the world may suffer from depression, which amounts to approximately 332 million people, affecting women more than men (6.9% versus 4.6%). If left untreated, this condition can lead to suicide, with an estimated 727,000 individuals who took their own lives in 2021. While complete remission is rare, Chand & Arif, suggest that at least 40% may achieve partial remission within a year thanks to available treatments.

Beyond its global footprint on human health, including the many lives that it can destroy, the economic burden of depression can have crippling effects on individuals, families, and society as a whole. A study by Greenberg et al., (2023) showed that adults with MDD were encumbered by a staggering $333.7 billion in 2019, or $16,854 per individual. According to the paper, 61.9% of this financial albatross ($206.4 billion) was associated with spendings unrelated to healthcare, “…including work-related costs of the adult with MDD from presenteeism ($43.3 billion; 13.0%), absenteeism ($38.4 billion; 11.5%), unemployment ($30.3 billion; 9.1%), all-cause mortality ($9.6 billion; 2.9%), and disability ($4.6 billion; 1.4%), and work-related costs of adults without MDD living in a household with an adult with MDD ($80.1 billion; 24.0% […])”

TREATMENT HORIZON

Conventional treatment with selective serotonin reuptake inhibitors (SSRIs) such as citalopram and fluoxetine have yield variable results. Premature discontinuation could account for a 17% increase in the relapse rate over a baseline of 39% during the course of one year (Duffy et al., 2021). Many other approaches are currently the focus of intensive investigation to improve general outcomes.

Cognitive behavioural therapy (CBT) is an objective-oriented method of identifying and changing negative thoughts. A recent meta-analysis of 409 trials involving 52,702 patients found that CBT can offer moderate to large improvements (g = 0.79) over control conditions (e.g., waitlist, care as usual), achieving 36% remission compared to only 15% in controls (Cuijpers et al., 2023).

Transcranial magnetic stimulation (TMS), which consists of non-invasively stimulating neurons in the superficial layers of the cortex, also shows promise, with a new study in 25 individuals demonstrating that accelerated repetitive TMS protocols can reach 52% response rates and 24% remission rates over three weeks (Motamed et al., 2025).

In a recent within-subject study of 120 patients with MDD, Ketamine infusion therapy was shown to have rapid and robust antidepressant effects, with noticeable reductions in suicidal ideation (Uyar & Güneş, 2025). Ketamine is a dissociative anesthetic drug that acts as a noncompetitive NMDA (N-methyl-D-aspartate) receptor antagonist and is widely used during emergency surgeries.

Electroconvulsive therapy (ECT), which induces transient seizures by sending small electric currents through the brain, remains one of the most effective interventions for treatment-resistant depression, with response rates superior to TMS (ECT: 52.8% versus TMS: 38.3%) and comparable to Ketamine (ECT: 41.2%-62.6% versus Ketamine: 46.3%-55.4%). Even so, ongoing concerns about cognitive side effects limit its wider applicability (Ruiz et al., 2025).

Psilocybin-assisted therapy (PAT) has emerged as a breakthrough approach. Psilocybin is a serotonergic hallucinogen found in “magic mushrooms,” and the pro-drug of the active psychotropic compound psilocin. A new meta-analysis of 16 studies found that while overall treatment effects were significant within relatively short- and long-term time windows (Cohen’s d = 1.69-2.10), the number of hours of adjunct psychotherapy showed no association with outcome measures. These results suggest that, unlike its compounding role in conventional antidepressants, the effects of psychotherapy in PAT appear to be superfluous, with observed reductions in depressive symptoms depending almost exclusively on the substance alone (Hultgren et al., 2025). Even so, PAT needs further research due to its unique challenges, such as biased participant selection and difficulties in maintaining study blinding (Hieronymus et al., 2025).

If you want to learn more about psychedelics, read/listen to my essay “Chaos - Manifesto of The Mind”:

While these considerations paint an overall positive picture for treatment prospects, showcasing multiple effective modalities that can be tailored to individual needs, the reality is that approximately two-thirds of people who need these therapies in high-income countries do not have access to them, raising important concerns about the future of global mental health.

DISTURBING THE FUNCTIONAL WEB

Resting-state functional magnetic resonance imaging (rs-fMRI) is a cornerstone of cutting-edge neuroimaging research. Initially believed to be noise, the spontaneous low-frequency fluctuations that rs-fMRI detects across the cortex constitute highly informative neural signatures of cognition, neurodevelopment, aging, and disease. With over 25,000 publications in peer-reviewed journals using this approach (Biswal & Uddin, 2025), rs-fMRI can tap into the functional architecture of the human brain while at rest (Yeo et al., 2011), revealing intrinsic modes of whole-brain organization known as resting state networks (RSNs). These are correlation-based clusters of temporal dynamics that capture associations in neural activity between different regions, and they include the default mode network (DMN), frontoparietal (FPN), ventral attention network (VAN), dorsal attention network (DAN), visual network (VN), somatomotor network (SMN), and limbic network (LN).

A new meta-analysis of 58 studies in which 4,518 total participants (e.g., 2,197 healthy controls and 2,321 patients with depression) had undergone rs-fMRI showed significant RSN alterations associated with depression (Zhang et al., 2025). The studies leveraged seed-based functional connectivity analysis, which looks at correlations between pre-determined brain areas with the rest of the cortex, and independent component analysis, a data-driven approach for decomposing complex signals into additive subcomponents.

This section breaks down the results from Zhang et al., focusing on functional alterations within each RSN (e.g., within-network changes) and between RSN pairs (e.g., between-network changes).

Within-network changes

• DMN showed both increased (hyper) and decreased (hypo) functional connectivity within its internal structure. The observed hyperconnectivity between the left precuneus and the rest of the DMN may indicate elevated self-referential processing, which is linked to depressive rumination. Alternatively, reduced connectivity between the bilateral angular gyrus and right middle temporal gyrus could point to compromised integration of external information as well as impaired cognition.

• FPN exhibited overall internal hyperconnectivity—namely, greater network integration of the the left caudate nucleus. Given the caudate’s role in processing pleasure/reward, this finding may indicate dynamic compensatory mechanisms aimed at attenuating the negative mental impact of depression.

Between-network changes

• DMN-FPN nodes showed increased connectivity, with consolidation of the right middle and inferior frontal gyri possibly pointing to uncontrollable self-directed negative ideations.

• DMN-SMN areas, such as the right superior temporal and Heschl’s gyri, showed weakened connectivity strength, suggesting suboptimal multimodal processing of sensorimotor inputs, which may affect the coordination between thought, sensation, and movement.

• LN-FPN connections weakened as LN-DMN links became stronger, which may denote an inability of executive brain areas to regulate heightened self-referential emotional processing.

• DAN-VAN associations exhibited both increased and decreased integration, speaking to complex disruptions of attentional networks and reflecting underlying difficulties in concentration. Hyperconnectivity, which is correlated with the duration of depression, may additionally serve as a neural biomarker for early diagnosis in susceptible individuals.

PUTTING ALL THE PIECES TOGETHER

Depression is a mood disorder with significant global prevalence, severe health-related consequences, and hefty budgetary drag. Conventional treatment with SSRIs have shown variable efficacy, with moderate success of cognitive behavioural therapy (CBT), and promising results from transcranial magnetic stimulation (TMS), ketamine infusion, electroconvulsive therapy (ECT), and psilocybin-assisted therapy (PAT), although access remains limited for two-thirds of those in need in high-income countries. Insights from neuroimaging studies using resting-state functional magnetic resonance imaging (rs-fMRI) reveal altered functional connectivity in depression across resting state networks (RSNs) like the default mode network (DMN) and frontoparietal network (FPN), with findings suggesting disruptions in self-referential processing, reward mechanisms, and attentional control, which may serve as biomarkers for diagnosis and treatment tailoring.

ESCAPING THE PULL OF THE 7^TH MAW

In Inferno, the first cantica of his tripartite Divine Comedy, Dante describes his journey into the hellscape of the underworld with his chaperone, Virgil. From crossing the river Acheron that lies between Earth and Hell to braving the frozen realm of Lucifer’s domain, Dante offers a detailed account of all the nine circles of damnation. Upon entering the Wood of the Suicides located in the seventh circle, Dante and Virgil come upon wailing trees with human faces and harpies flying overhead. Following suit with his guide’s instructions, Dante snaps off a branch from a nearby thornbush, at which the trunk began to speak. After a brief exchange, the trunk explains the predicament of the wretched souls who end up there:

When the savage spirit quits the body from which it has torn itself, then Minos sends it to the seventh maw. If falls into the wood, and there’s no place to which it is allotted, but wherever fortune has flung that soul, that is the space where, even as a grain of spelt, it sprouts. It rises as a sapling, a wild plant; and then the Harpies, feeding on its leaves, cause pain and for that pain provide a vent. Like other souls, we shall seek out the flesh that we have left, but none of us shall wear it; it is not right for any man to have what he himself has cast aside. We’ll drag our bodies here; they’ll hang in this sad wood, each on the stump of its vindictive shade.

When the pieces that make up the functional architecture of your brain no longer fit together the way they used to, you may indeed find yourself subjected to the gravity of the seventh circle. As someone who knows, firsthand, just how much confusion and heartache losing a loved one to suicide can bring to those mourning, the passage above left me with a lot to reflect upon while reading Dante’s cantos. Though there is little in it to convince me that what it describes is indeed the cosmic lot of individuals who take their own lives, it does serve as a window into the zeitgeist of 14^th century Europe regarding felo de se. The truth is that life is full of hurdles, and while it may disdainfully serve us a gloomy portion of vicissitudes every so often, it does not always grace us with the right mixture of good genes and good upbringing to countenance it. But as long as you keep trying to put the pieces of your life’s puzzle back together, you may find that you always have something worth looking forward to.

If you enjoyed this article, you might also be interested in my essay titled “Addiction (Substance Use Disorder) - The brain’s internal odyssey”:

Share

Subscribe now

Follow me on X

Leave a comment

Add me on LinkedIn

Share ZAGROSCIENCE

Check out ZAGROSCIENCE on YouTube

REFERENCES

Biswal, B. B., & Uddin, L. Q. (2025). The history and future of resting-state functional magnetic resonance imaging. Nature, 641(8065), 1121-1131.

Chand, S. P., & Arif, H. (2025). Depression. StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK430847/

Cuijpers, P., Miguel, C., Harrer, M., Plessen, C. Y., Ciharova, M., Ebert, D., & Karyotaki, E. (2023). Cognitive behavior therapy vs. control conditions, other psychotherapies, pharmacotherapies and combined treatment for depression: A comprehensive meta‐analysis including 409 trials with 52,702 patients. World Psychiatry, 22(1), 105-115.

Dante Alighieri. (1982). The divine comedy of Dante Alighieri: Inferno: A verse translation (A. Mandelbaum, Trans.). Bantam Books. (Original work published ca. 1314)

Duffy, L., Clarke, C. S., Lewis, G., Marston, L., Freemantle, N., Gilbody, S., Hunter, R., Kendrick, T., Kessler, D., King, M., Lanham, P., Mangin, D., Moore, M., Nazareth, I., Wiles, N., Bacon, F., Bird, M., Brabyn, S., Burns, A., Donkor, Y., … Lewis, G. (2021). Antidepressant medication to prevent depression relapse in primary care: the ANTLER RCT. Health technology assessment (Winchester, England), 25(69), 1–62. https://doi.org/10.3310/hta25690

Greenberg, P., Chitnis, A., Louie, D., Suthoff, E., Chen, S. Y., Maitland, J., Gagnon-Sanschagrin, P., Fournier, A. A., & Kessler, R. C. (2023). The Economic Burden of Adults with Major Depressive Disorder in the United States (2019). Advances in therapy, 40(10), 4460–4479. https://doi.org/10.1007/s12325-023-02622-x

Hieronymus, F., López, E., Sjögren, H. W., & Lundberg, J. (2025). Control group outcomes in trials of psilocybin, SSRIs, or esketamine for depression: a meta-analysis. JAMA Network Open, 8(7), e2524119-e2524119.

Hultgren, J., Hafsteinsson, M. H., & Brulin, J. G. (2025). A dose of therapy with psilocybin–A meta-analysis of the relationship between the amount of therapy hours and treatment outcomes in psychedelic-assisted therapy. General Hospital Psychiatry.

Motamed, M., Fathi, A., Hajikarim-Hamedani, A., & Alaghband-Rad, J. (2025). Accelerated repetitive transcranial magnetic stimulation in the treatment of depressive disorder resistant to a course of antidepressant medication. Annals of general psychiatry, 24(1), 30. https://doi.org/10.1186/s12991-025-00572-y

Ruiz, A. C., Haseeb, A., Baumgartner, W., Leung, E., Scaini, G., & Quevedo, J. (2025). New insights into the mechanisms of electroconvulsive therapy in treatment-resistant depression. Frontiers in psychiatry, 16, 1614076. https://doi.org/10.3389/fpsyt.2025.1614076

Uyar, B., & Güneş, M. (2025). An evaluation of the efficacy and side effects of a single dose of ketamine in major depressive disorder. Postgraduate medical journal, 101(1198), 716–722. https://doi.org/10.1093/postmj/qgae142

World Health Organization. (2025, August 29). Depressive disorder (depression). World Health Organization. https://www.who.int/news-room/fact-sheets/detail/depression

Yeo, B. T., Krienen, F. M., Sepulcre, J., Sabuncu, M. R., Lashkari, D., Hollinshead, M., Roffman, J. L., Smoller, J. W., Zöllei, L., Polimeni, J. R., Fischl, B., Liu, H., & Buckner, R. L. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of neurophysiology, 106(3), 1125–1165. https://doi.org/10.1152/jn.00338.2011

Zhang, Z., Zhang, Y., Wang, H., Lei, M., Jiang, Y., Xiong, D., Chen, Y., Zhang, Y., Zhao, G., Wang, Y., Zhang, W., Xu, J., Zhai, Y., An, Q., Li, S., Hao, X., & Liu, F. (2025). Resting-state network alterations in depression: a comprehensive meta-analysis of functional connectivity. Psychological medicine, 55, e63. https://doi.org/10.1017/S0033291725000303

Over and over again men are blinded by too violent motivations and too intense frustrations into blind and unintelligent and in the end desperately dangerous hates of outsiders. And the expression of these their displaced hates ranges all the way from discrimination against minorities to world conflagrations. What in the name of Heaven and Psychology can we do about it? My only answer is to preach again the virtues of reason—of, that is, broad cognitive maps.

~ Edward Tolman — Cognitive maps in rats and men (1948)

Why should you subscribe to ZAGROSCIENCE?

If you’re interested in hardcore neuroscience with an artistic flair, then you’ll enjoy ZAGROSCIENCE, which comprises an expanding library that includes audiovisual and written/narrated pieces on memory, consciousness, psychedelics, epilepsy, addiction, and much more. I’m also launching a new podcast series called The Abstract, where I’ll be interviewing leading experts in the field of brain research—some of whom are professional colleagues of mine. Valued patrons of ZAGROSCIENCE will be able to submit questions in written, audio, or video format, which I will curate and share on the show with my guests.

Why should you care about what I have to say about neuroscience?

I’m a cognitive neuroscientist who obtained his PhD at McGill University in 2024, specializing in memory and consciousness, with years of research experience at the Montreal Neurological Institute (aka, The Neuro) and over 25 (and counting) peer-reviewed publications in top scientific journals (i.e., Nature Neuroscience, Cerebral Cortex, Brain, Neuroimage, etc.). Here’s a PubMed link to my scholarly work (Tavakol is the surname under which I publish):

https://pubmed.ncbi.nlm.nih.gov/?term=shahin%20tavakol%5BAuthor%5D&sort=fauth

On top of all that, I like to think that my writing is clear and engaging as I enjoy mixing in relevant themes from disciplines other than neuroscience (i.e., philosophy, literature, history, mythology, etc.).

Why should you read this particular article?

Factorization, conjunction, and path integration are believed to underlie how your brain makes sense of the world. So by understanding how these processes unfold inside your brain, the world around you might just become a little more sensible.

Subscribe now

WHAT IS FACTORIZATION?

Consider the mathematical expression:

Since A is the common factor between AB and AC, we can “pull it out”:

This process, known as factorization, retains the underlying information, such that:

So why bother rewriting an expression in its factorized form if the content stays the same? The truth is, it adds information about information (or meta-information, if you prefer). Far from being a mathematical exercise in futility, factorization is a way of simplifying expressions, making it possible to extract general patterns across complex formulae that reduce complexity and streamline arithmetic computation.

Related mathematical and geometric operations are believed to be at the core of automatic mental processes that shape our lives. In the current issue of ZAGROSCIENCE, we will explore a prominent computational model of human cognition known as the Tolman-Eichenbaum Machine, which replicates three key operational features of the hippocampal-entorhinal system: factorization, conjunction, and path integration.

WHERE DOES THE NAME TOLMAN-EICHENBAUM COME FROM?

In 1948, Edward Tolman conducted two rodent experiments that ushered in the modern age of neuroscience. In the first one, rats trained to reach a food box inside a maze adapted their path finding behaviour when the contraption was altered, suggesting that they were using a mental representation of their environment for navigation. In the second experiment, rats trained in a T-maze to go left or right continued to follow the same direction even when the set was flipped, indicating that food target was not the only factor responsible for wayfinding. Taken together, Tolman argued that his findings showed that rodents can form a mental map of their surroundings, which he dubbed the cognitive map.

Many years later in 1993, Howard Eichenbaum and another colleague proposed the relational memory theory, which described the functional role of the hippocampus and associated areas as a relational system that binds multisensory information. This concept led Eichenbaum to argue that previously established hippocampal place cells, which were shown to fire in response to unique locations in a given environment, should instead be rebranded as relational cells, given their involvement in a wider scope of associative cognitive processes.

The Tolman-Eichenbaum Machine, thus, borrows its name from these two pioneers who advanced our understanding of spatial/relational brain-behaviour mechanisms.

Note: For more information on Tolman, Eichenbaum, cognitive maps, relational memory, and my own contributions to the field, see my inaugural essay on titled Of Mice and Men - A Primer on the Neuroscience of Memory:

WHAT IS THE TOLMAN-EICHENBAUM MACHINE?

Proposed by Whittington et al., (2020), the Tolman-Eichenbaum Machine (henceforth, TEM) is a mechanistic framework that explains your brain’s capacity to deconstruct, store, and manipulate the complex associations between the constituent elements of consciousness. At its core, TEM is modelled on the specialized neurons inside the hippocampus and entorhinal cortex that allow us to flexibly interact with and navigate the physical, social, and abstract domains of our lives.

WHAT IS THE HIPPOCAMPAL-ENTORHINAL SYSTEM

The hippocampus (from Greek hippos ‘horse’ + kampos ‘sea monster’) is a seahorse-shaped structure within the three-layered allocortex inside the medial temporal lobe (MTL) known to play a major role in memory, emotion, and general cognition. It is composed of several subregional structures known as subfields, including the dentate gyrus (DG), cornu ammonis (CA) 1-4, and subiculum.

The entorhinal cortex (EC)—also part of the MTL—is immediately adjacent to the hippocampus, but situated in the six-layered isocortex (or neocortex). It represents the primary neocortical source of input/output to the hippocampus. Together, these structures form the hippocampal-entorhinal system, a neural complex that sustains relational memory across multiple sensory modalities.

WHAT SPECIALIZED CELLS ARE INVOLVED IN TEM?

• Place cells (or relational cells) are hippocampal neurons that fire in response to a specific spatial location, called the place field. These were originally observed in rats with direct electrophysiological recordings of the dorsal CA1 subfield (O’Keefe & Dostrovsky, 1971) and later validated in humans using intracranial electrodes and virtual reality (VR) in patients with drug-resistant epilepsy (Ekstrom et al., 2003).

• Grid cells are found in the medial EC and are functionally similar to hippocampal place cells. However, while the hippocampal place field of a given place cell is confined to a single location in the immediate surroundings, the entorhinal grid field is instead marked by evenly spaced points that form a tessellating triangular pattern. First described in rodent experiments employing direct electrode recordings (Hafting et al., 2005), grid cells were later confirmed in drug-resistant epilepsy patients assessed with intracranial recordings while immersed in VR (Jacobs et al., 2013). Grid cells are also involved in path integration, which we will discuss later.

• Border cells are sparsely distributed neurons in the medial EC that spike in response to proximal boundaries, such as walls and barriers. Initially recorded in freely moving rats navigating different enclosures, border cells have been proposed to anchor place and grid fields to the local environment by delineating its bounds (Solstad et al., 2008).

• Object-vector cells, also found in the medial EC of foraging rodents, are tuned to the distance and direction of objects regardless of the rat’s orientation or the size/composition of the item (Høydal et al., 2019). Object-vector cells offer a cellular mechanism for mapping position in relation to physical landmarks.

• cells in the lateral EC, unlike their medial EC counterparts, are responsive to non-spatial information. It has been shown in rodents that lateral EC cells play a role in the multisensory processing of objects, such as toys that vary in texture, shape, colour, and size (Deshmukh et al., 2011).

Note: Drug-resistant epilepsy (DRE) is arguably the most well-established human model of memory impairment, serving as the focus of countless studies since the 1950s. Given that intracranial recordings are used to monitor disease progression in patients, DRE offers a unique opportunity to directly tap into the neural substrates of various forms of cognition, which explains the aforementioned discoveries of place and grid cells in humans. Read my article Cortical Wildfires for more information on this fascinating topic:

HOW DO TEM CELLS PROCESS THE CONTENTS OF CONSCIOUSNESS?

A conscious experience can be construed as the product of two primary components: abstract associations between nondescript items and identifiable characteristics that make each item its own unique entity. Think, for example, of a family gathering. In this scenario, associations can be thought of as how each individual is related to everyone else while identifiable characteristics can be the red ribbon little Suzie is wearing around her hair or how grandpa is always clearing his throat when he is about to talk. Each circumstance comes with its own set of structural knowledge (i.e., family bonds) and sensory representations (i.e., identifiable features for each person).

According to the paper by Whittington et al., TEM posits that the specialized neurons of the hippocampal-entorhinal complex can factorize consciousness by extracting its underlying structural knowledge (via medial EC cells) and sensory information (via lateral EC cells). In a process known as conjunction, the hippocampus, which receives its main inputs from medial/lateral EC, can bind both components, forming what is called a conjunctive code, which can be stored in memory. This mechanism is key for learning via generalization, as conjunctive codes can be recalled and factorized anew to reveal latent structural knowledge that can be recombined with newly processed multisensory data.

As I have written in a previous work [https://zagroscience.substack.com/p/the-rabbit-marching-up-your-sleeve], your brain is constantly trying to make sense of the present by rehashing what it has previously learned—and sometimes, the clash between the two can play odd tricks on your mind…

WHAT ABOUT PATH INTEGRATION…WHAT IS THAT?

Let us assume that you live in a simple 2D Cartesian space at the origin (0,0). You move 3 paces right along the x-axis. Your current coordinates are (3,0). You then move 4 paces up along the y-axis. Your new location is (3,4). Path integration is knowing that to get back to the origin in one straight line, you must traverse 5 paces at 126.87° clockwise. From a cognitive neuroscience perspective, path integration is an incidental functional feature of the hippocampal-entorhinal system to implicitly compute the opposite of the resultant displacement vector (i.e., 5 paces at 126.87° clockwise) based on the individual component vectors (i.e., 3 paces right along the x-axis + 4 paces up along the y-axis). As you will find out, your brain can apply path integration to solve relational problems beyond the scope of spatial navigation…

WHAT DID WHITTINGTON ET AL., (2020) FIND?

The authors developed an artificial neural network modelled after the principles of information factorization, conjunction, and path integration. They trained and tested their model on a number of graphs with different structural and idiosyncratic properties, where the artificial agent predicted the next item in a sequence of probabilistic transitions between nodes [note: in graph theory, a graph is a network comprised of nodes connected via edges]. Results from simulations showed that their machine learning algorithm reached 100% accuracy in item prediction either by sampling all nodes alone (i.e., node agent) or all nodes and edges together (i.e., edge agent). Interestingly, the node agent achieved peak performance much quicker than its counterpart. Consider this figure from Whittington et al.:

What the figure above shows is how a node agent that visits all nodes—but only a fraction of all edges (i.e., 18/42 or ~43%)—can learn the structural scaffolding of the graph and attain perfect prediction accuracy in a fraction of the time (measured here in network steps) than an edge node that must sample all connections.

For the node agent, path integration is how the model learns to fill in the missing links (i.e., edges not sampled). What is noteworthy is that the algorithm can perform path integration regardless of whether the graph represents movements in physical space or conceptual associations in abstract space. In fact, so long as there is an intrinsic relational structure to it, the graph can schematize just about any feature space. Take a look at the following figures from the same paper:

The figure above displays three different relational schemes that showcase transitive (A), social (B), and 2D directional (C) connections. Given enough trials, the node agent will learn the latent structure of each graph and perform path integration to infer previously unseen edges. For instance, the associations in (A) show that the motorcycle is faster than the bicycle, which is faster than the scooter. Learning these links enables the model to determine that the motorcycle is faster than the scooter, even though it has not explicitly encoded this information during training. Similarly in (B), learning that Bob is Alice’s brother, and Alice is Jane’s mother, allows the algorithm to infer that Jane is Bob’s niece. Regardless of the nature of relations (i.e., transitive, social, or directional) or the identity of node objects (i.e., modes of transportation, humans, or landmark structures), the TEM neural network can flexibly manipulate information through factorization, conjunction, and path integration to learn novelty, recall associations, and merge old blueprints with new sensory stimuli, driving flexible behaviours that eventually meet the standards of peak performance.

WHAT IS THE TAKE-HOME MESSAGE?

The Tolman-Eichenbaum Machine (TEM), proposed by Whittington et al., (2020) and named after two pioneers who advanced our understanding of spatial/relational mechanisms, is a mechanistic model designed to explain your brain’s capacity to decompose, store, and manipulate the complex associations that constitute the contents of consciousness. This paradigm is based on the specialized neurons within the hippocampal-entorhinal system, an integrated neural complex that sustains relational memory across various sensory modalities. The system is believed to employ factorization to extract underlying structural knowledge—via medial EC cells—and sensory information—via lateral EC cells—from experience. The hippocampus then performs conjunction to bind these components into a storable conjunctive code. Furthermore, experiments utilizing an artificial neural network modelled on TEM demonstrated how path integration can infer previously unseen connections, enabling the TEM algorithm to flexibly manipulate information regardless of whether the context involves spatial or nonspatial associations.

ONE LAST QUESTION: WHERE IS THE GHOST IN THE MACHINE?

Contemporary neuroscience has little to say—if anything at all—about a mind that operates independently of the brain. As a physical science of the latter, it is, ultimately, bound to the atoms that make up the molecules that make up the neurons that make up the 3 pounds of tapioca pudding inside your skull. To the extent that the machine may be possessed, neuroscience can be regarded as the empirical framework that exorcises it, demystifying human nature in the process. It is said that nature, in general, reveals herself through the language of mathematics. The ghost in the Tolman-Eichenbaum Machine may just be, then, the sum-totality of mathematical operations inside the temporal lobes. And so, if human nature is reducible to the inner workings of your brain, then it stands to reason that these mechanisms spell out the universe.

Note: For a survey of modern scientific models of consciousness, reductionist vs. nonreductionist schools of thought, contemporary experimental paradigms, and the implications of memory as the basis for consciousness, I invite you to read the two essays I’ve listed below.

From Atoms to Qualia - The Neuroscience of Consciousness:

The rabbit marching up your sleeve - Mnemonic processing as an exegesis of consciousness:

FOLLOW THE RABBIT HOLE AND PICK YOUR POISON(S) FROM THIS LIST:

Share

Subscribe now

Follow me on X

Leave a comment

Add me on LinkedIn

Share ZAGROSCIENCE

Check out ZAGROSCIENCE on YouTube

REFERENCES

Cohen, N. J., & Eichenbaum, H. (1993). Memory, amnesia, and the hippocampal system. MIT press.

Deshmukh, S. S., & Knierim, J. J. (2011). Representation of non-spatial and spatial information in the lateral entorhinal cortex. Frontiers in behavioral neuroscience, 5, 69.

Duvernoy, H. M., Cattin, F., & Risold, P-Y. (2013). The human hippocampus: functional anatomy, vascularization and serial sections with MRI (4th ed.). Berlin, Heidelberg: Springer-Verlag

Eichenbaum, H., & Cohen, N. J. (2001). From Conditioning to Conscious Recollection: Memory Systems of the Brain. Oxford University Press.

Ekstrom, A. D., Kahana, M. J., Caplan, J. B., Fields, T. A., Isham, E. A., Newman, E. L., & Fried, I. (2003). Cellular networks underlying human spatial navigation. Nature, 425(6954), 184-188.

Hafting, T., Fyhn, M., Molden, S., Moser, M.-B., & Moser, E. I. (2005). Microstructure of a spatial map in the entorhinal cortex. Nature, 436(7052), 801-806.

Høydal, Ø. A., Skytøen, E. R., Andersson, S. O., Moser, M. B., & Moser, E. I. (2019). Object-vector coding in the medial entorhinal cortex. Nature, 568(7752), 400-404.

Jacobs, J., Weidemann, C. T., Miller, J. F., Solway, A., Burke, J. F., Wei, X.-X., Suthana, N., Sperling, M. R., Sharan, A. D., & Fried, I. (2013). Direct recordings of grid-like neuronal activity in human spatial navigation. Nature Neuroscience, 16(9), 1188-1190.

O’Keefe, J., & Dostrovsky, J. (1971). The hippocampus as a spatial map: preliminary evidence from unit activity in the freely-moving rat. Brain research.

Solstad, T., Boccara, C. N., Kropff, E., Moser, M. B., & Moser, E. I. (2008). Representation of geometric borders in the entorhinal cortex. Science, 322(5909), 1865-1868.

Tavakol, S., Li, Q., Royer, J., Vos de Wael, R., Larivière, S., Lowe, A., ... & Bernhardt, B. (2021). A structure–function substrate of memory for spatial configurations in medial and lateral temporal cortices. Cerebral Cortex, 31(7), 3213-3225.

Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological review, 55(4), 189.

Whittington, J. C. R., Muller, T. H., Mark, S., Chen, G., Barry, C., Burgess, N., & Behrens, T. E. J. (2020). The Tolman-Eichenbaum machine: unifying space and relational memory through generalization in the hippocampal formation. Cell, 183(5), 1249-1263.