Bitwise

by David Auerbach

  I went along with the color-coding at the corporate retreat. I put on a green shirt and accepted my lot, but when it came time to form teams of five or six people for competitive activities (build the tallest tower out of office supplies, give accurate navigation instructions to a blindfolded teammate, etc.), I chose not to group with the other greens. I ended up joining a group of mostly blues—the friendly harmonious ones. It was just a corporate retreat; I knew most greens would overanalyze everything, like pseudoscientific personality mapping. The blue people were kinder and more relaxed overall, and I enjoyed myself. Maybe deep inside I’m blue. But the database that maps me as a green, an INTJ, an MMPI-2 depressive, or any one of a million other categories wouldn’t know it.

  *1 Or, more precisely, mathematical abstractions, whether numbers or bits or sets or functions or spatial figures or what have you. These all fall under the rubric of data, but “data” has become a hopelessly vague term that goes far beyond the reach of computers, so I will tend to say “numbers,” the underlying representation for (most) computer data.

  *2 The debate around transmedicalism is only one example of how unresolved these questions are.

  *3 The studies claiming that left-handed people die earlier than right-handed people are mostly bunk. One study showed that they died younger only because they couldn’t find any left-handed people above a certain age. This, it turned out, was because they had all been forced to be right-handed in school during the relevant time period.

  *4 Chinese civilization and culture hold a record for stability and continuity that puts any other large region of the world to shame. It’s a testament to the robustness of Confucian thinking that it remains standing in somewhat recognizable form after almost 2,500 years. As a cultural entity, Christianity probably comes closest for sheer duration, but the multitude of belief systems and cultural movements called “Christianity” and the varieties of ideas and practices attributed to the name are so wildly disparate that you’d be hard-pressed to identify any commonality between them short of their veneration of a person whose birthdate (approximately) sets the inflection point of the world’s dominant calendar.

  *5 Traditionally, almost any folk dichotomy sooner or later comes to echo the primeval dichotomies of male and female, light and dark, and life and death. These are fairly loaded dichotomies that are best avoided when one is seeking a supposedly neutral typology.

  *6 Tripartite schemata are fascinating. They discourage opposition and symmetry, while encouraging instability and conflict. Things are frequently divided into three, whether Plato’s tripartite soul or the Christian Trinity or Freud’s psyche, but things are far less often classified into one of three separate categories.

  *7 Anecdotally, Chinese culture seems to have a fondness for fivefold structures: Confucius’s five bonds, the five elements of Wu Xing, and the five elements of qi in Chinese medicine. East Asian cultures hold a great deal of numerological superstition, particularly around the number four (which is homophonous with the word for death), so perhaps this has served as a disincentive against fourfold structures. Declaring such a low number to be unlucky is a tremendous inconvenience; far more buildings have fourth floors than thirteenth floors. Better to make large numbers unlucky: I can’t think of too many occasions where superstitions around the number 666 have come into play. The prize for inconvenience must go to the Russian superstition that and even numbers of flowers are unlucky, forcing purchasers to count buds lest they give the recipient an even number.

  *8 George Makari’s Revolution in Mind: The Creation of Psychoanalysis paints an unflattering portrait of the dawn of psychoanalysis, showing its pioneers to be ingenious but often undisciplined creative minds grasping in the dark toward the incomprehensibilities of the psyche, trading creation myths among one another in a farcical race to create a unified theory of the mind ab ovo. The result was a secular mythology of the mind whose greater contours remain with us today. Sigmund Freud, for all his intellectual caprice, still comes off as the sharpest mind of the lot, Alfred Adler as the most sensible. One psychoanalytic theorist was far more aggressively specious, racist, and arrogant than the others: Carl Jung.

  *9 In the sidereal, astronomically correct zodiac, my birthdate falls under the short thirteenth sign of the Serpent Holder Ophiuchus, progenitor of the archetypal Greek healer Asclepius (linked to the Egyptian pharaoh and healer Imhotep), a sign that apparently classifies me as a secretive mastermind with grand architectural plans and a distaste for authority and injustice, prone to being hypercritical, alienated, and impatient. I think astrology is nonsense, but as far as signs go, I could have done a lot worse. I also was born in the Chinese year of the Dragon—the arrogant yet humane Fire Dragon, specifically—and it’s a shame these not-inappropriate signs should have been wasted on someone who doesn’t believe in them.

  *10 Calvino was a member of the Oulipo, and this was one of his experiments in writing stories based around formal constraints, much like Queneau’s Exercises in Style and Perec’s Life: A User’s Manual.

  *11 There is an old Simpsons episode in which a Tarot reader explains that Death signifies beneficial change, but gasps in horror when The Happy Squirrel appears.

  5

  SELF-APPROXIMATIONS

  The Big Five (or Six)

  Putting facts into nice cleanly demarcated buckets of explanation has its advantages—for example, it can help you remember facts better. But it can wreak havoc on your ability to think about those facts.

  —ROBERT M. SAPOLSKY

  MYERS-BRIGGS IS RECREATIONAL. The MMPI is clinical. What if there was a model that claimed to capture the entirety of a person’s personality, not just a few aspects of it? There already are several. Somewhere in between the MMPI and Myers-Briggs are popular yet scholarship-backed taxonomies like the “Big Five” personality traits. The “Big Five” has achieved prominence over the past twenty-five years, after a number of psychology researchers reached consensus on the general model in the 1980s. The model clusters personality traits under five umbrella terms: Neuroticism, Agreeableness, Conscientiousness, Extroversion, and Openness (which produce the acronyms OCEAN or CANOE, depending on one’s preference for seeing the psyche as a murky depth or a leaky boat). Each is a linear spectrum.

  The best-known variant of the family of Big Five models is the NEO Five-Factor Inventory, created by psychologists Paul Costa and Robert McCrae.*1 Costa and McCrae claim the Big Five model is cross-culturally universal and reflects basic dispositional tendencies that are predominantly fixed by genetics at birth, though both claims have been disputed. In his popular introduction Personality, Daniel Nettle deems the Big Five to be the master model of personality: “Most constructs that had previously been measured can actually be subsumed under the five-factor framework….It is very likely [the Big Five] will turn out to be shorthand for suites of differences in neural structure and function across multiple brain regions.” This audacious claim—that we have discovered the neurobiological structure of human personality—has also been disputed.

  The Big Five personality traits

  Two other psychologists, Kibeom Lee and Michael C. Ashton, have created a six-factor model, HEXACO, that takes OCEAN’s Big Five and adds a sixth, purportedly independent axis: Honesty vs. Humility. In this scheme, the Big Five’s Neuroticism has been renamed Emotionality, while Openness is also labeled Intellect/Imagination/Unconventionality. (Extroversion becomes the X in HEXACO.) The authors of Big Five systems are, unsurprisingly, dismissive of HEXACO. McCrae downplays Honesty-Humility as a superfluous correlate of Agreeableness.*2

  The six-factor HEXACO personality model

  Personality models like the Big Five and HEXACO promise neat, authoritative classifications of our personalities, more scientific than Myers-Briggs and broader than the MMPI. Both are well suited to computational classification. The problem is which model to pick. McCrae and Cost
a boldly say, “The existence of these five factors is simply an empirical fact, like the fact that there are seven continents on the earth,” but scientific rigor does not permit that conclusion, any more than we can say that the four humors are an empirical fact. Such things are heuristic constructs. With the Big Five and HEXACO, we’ve been given two yardsticks, each with different sets of markings on it, and asked to decide which yardstick has the correct markings. There is no correct answer. We are not measuring an objective phenomenon, but rather trying to measure the subjective measures themselves. Each model draws attention to certain personality distinctions while overshadowing distinctions that the model cannot accommodate. The personality models, while not arbitrary, are selective and inadequate. OCEAN and HEXACO proponents each point to behavioral studies supporting their respective views, but until we are able to determine personality from a person’s genes and neurobiology, science cannot choose one over the other. Yet the dispute will be “resolved” when one model proves more popular than the other—or, more likely, when some third model supplants both.

  What if our OCEAN (or HEXACO) types were associated with our Facebook profiles, our corporate HR records, our government files, and so on? In fact, OCEAN and HEXACO are used by corporate training organizations, just as Myers-Briggs typologies are. Would we want to be placed into a HR database indicating we’ve scored low on Agreeableness? Once we start using one system, like OCEAN, there are high costs to transitioning to another like HEXACO. Faced with the burden of having to reclassify people, there will be a strong incentive to believe that the existing classification is indeed the correct one.

  In practice, we need taxonomies to provide some coarse classification of types. Humans speak constantly in reductive generalizations, because we lack the physical and mental capacities to process the world in all its raw detail. The danger lies in adopting and privileging a single taxonomy. When we do that, we run the risk of stomping on the complexities that slip through the sieve of whichever conceptual system we use. And the more we hew to one system, the less likely we are to revise that system or transition to a new and possibly better one. No four-, five-, or six-factor system will ever be adequate for the complexities of human psychology, but a new system is often preferable to the existing one, if only because it preserves variety and reminds us that our current system is hardly the be-all and end-all.

  Diagnostics and Statistics

  It would be indeed unusual if it turned out that the set of orders that our mind is able to construct and accept, having as it does a deep sense of “understanding the essence of things,” matches precisely the set of all possible orders to be detected in the Universe as a whole. We should admit that this is not impossible, yet it does seem highly improbable.

  —STANISLAW LEM

  MY PARENTS had three huge bookshelves along the wall of our living room, and as I grew up I made my way through their collection. There was a good dose of golden-age science fiction, Carl Barks’s Donald Duck comics, the World Book Encyclopedia, Max Shulman’s Dobie Gillis stories, and Martin Gardner’s annotated Lewis Carroll. There were also the mid-century books for adults I didn’t read until much later: Serpico, Catch-22, and Moss Hart’s Act One, alongside psychiatric texts like Freud’s Totem and Taboo, a title I found creepy. But the most perplexing book they had was one I also spent a great deal of time with: the Diagnostic and Statistical Manual of Mental Disorders, 3rd Edition, Revised (the DSM-III-R for short), the psychiatrist’s bible, published by the American Psychiatric Association. This 1987 reference book, which was over five hundred pages of dense type, contained an elaborate numerical taxonomy of every mental disorder under the sun, from alcoholism to schizophrenia to anorexia, along with detailed criteria for diagnosing whether a patient possessed each disorder. I say every mental disorder because, as I learned, psychiatric practice often required that all patients be labeled with one or more of the labels in the DSM.

  Both of my parents are psychiatrists, and the 1980s were an exciting and transformative time for psychiatry for two main reasons: the introduction of the new wave of SSRI antidepressants such as Prozac and Zoloft, and the diagnostic revolution of the DSM-III (1980) and its successor, the DSM-III-R (1987). As a child, I witnessed two manifestations of these revolutions in our house: the preponderance of Prozac- and Zoloft-branded paraphernalia like pens, mousepads, Post-it note cubes, and penlights, and the constant presence of the intense blue cover of the DSM-III-R.

  By the late eighties, my family owned an IBM-compatible computer (a Compaq, specifically) as well as an Epson dot-matrix printer.*3 My parents were not good with computers, so they enlisted me to help format, spell-check, and print their psychiatric reports for them. We used WordPerfect 5.1, a solid if austere word processing program that presented little more than a blank blue screen on which to type, along with a set of function key shortcuts that led to menus and other various options.*4 The blue screen matched the blue of the cover of the DSM-III-R.

  My father worked for the county health system but occasionally did consultation work for the state in assessing disability candidates. The state provided no information on the candidates, other than to let him interview them a single time to provide input into the state’s determination of whether or not the candidate was entitled to disability benefits. Given this suboptimal clinical situation, my father tried to be charitable, if cautiously so. The state was not inclined to be charitable. The reports couldn’t be conclusive in any way under the circumstances, but the state required them before they would pay disability. And a big part of the report was the DSM diagnosis.

  My father typed his reports, his typewriter-trained fingers loudly hammering the keys, then, after blanking all identifying information, he asked me to work my computer magic. I would then underline and bold the appropriate parts, align everything properly, leave room for his signature, and print out the final product. These consultation reports contained summaries of the patients’ backgrounds, complaints, and cognitive assessments, such as whether they could recite the alphabet forward and backward, spell simple words correctly, name the first and last five presidents, etc. At the end was the terse Prognosis, which was frequently “good” or “fair” but sometimes “guarded” (which, as my father explained to me, was the opposite of “good”). In between the summary and the prognosis was the all-important diagnosis, consisting of five “axes,” as defined by the DSM. Here they are, roughly, before they were eliminated in DSM-5:

  Axis I: The diagnostic code for the principal disorder, as defined by the DSM.

  Axis II: Personality disorders and intellectual disabilities.

  Axis III: Relevant non-psychiatric medical conditions.

  Axis IV: Relevant psychosocial, environmental, and/or situational factors.

  Axis V: Global Assessment of Functioning (GAF) score.

  The first and last were by far the most important, to the extent that my father often left the middle axes blank. This was the first time I’d witnessed human behavior classified in such a concrete and discrete fashion. As a budding quant, I was fascinated by the strangeness of it. People I did not know were being described in a clinical language that didn’t match anything of my understanding of people. This clinical language also seemed to describe a private world to which I didn’t have access. I learned words like “anxiety” and “borderline” and “obsessive-compulsive,”*5 but none of them appeared to apply to human behavior as I witnessed it. It was a foreign dialect.

  Axis V, the Global Assessment of Functioning, seemed simplistic to me even then. It was a scale from 1 to 100, where 100 was “superior functioning,” 50 was “serious impairment in functioning,” and anything under 20 signified danger to one’s self or others. The GAF struck me as a bit…general. But then, that appeared to be its role: a rough measure of how badly off patients were and thus how in need of treatment they were. I asked my father how he
could possibly differentiate between a 55 and a 56, and he told me that he couldn’t, and he didn’t. No one could.

  Axis I was much more complicated and much more interesting. It was an elaborate list of coded diagnoses, usually identified by three digits and two decimals, which elaborated every possible psychiatric diagnosis that could go into a patient’s file. Post-traumatic stress disorder, bipolar, paranoid schizophrenia, borderline personality disorder; just as these terms had been brought into common parlance in the 1980s through usage of the DSM-III, they became known to me through my parents’ use of them as codes. It was their definitions that were odd.

  Every disorder had a number, but the numbers were misleading. I had read Martin Gardner’s polemics against pseudosciences like Scientology and Velikovskyism, and I was on the lookout for more. I liked math, but the DSM wasn’t math. It was just numbers. Catatonic schizophrenia did not become a more objective label simply by assigning it the number 295.2. The DSM notoriously uses a criterial system to determine diagnoses: one need not possess all the markers of a disorder to be diagnosed with it, only enough of them. Many disorders entail having “5 of the following 7” or “at least 4 of the following 10” behaviors or tendencies listed.