In defense of prescriptive labeling

[Content note: I hold my main conclusion with a moderately low level of confidence, and I’m open to being convinced otherwise.  Content heavily involves feminists and nerds, but the reader can rest assured that I’ll steer clear of discussing the interactions between them.]

Here I am, with a blog dedicated to exploring the commonly-encountered misunderstandings that lead to disagreements, and somehow I still haven’t gotten around to describing my views on prescriptive versus descriptive definitions for categories of people.  Since I see this as kind of vital to lot of current conversations, I think it’s about time I wrote about this.

I’ll start by making my terms clear.  (I’m not entirely certain that this will be the most correct use of the terms, but this is the way I think of them.)  A prescriptive definition is one which describes a category in terms of predetermined characteristics that one believes members of that category should hold to.  A descriptive definition is one which describes a category according to the shared characteristics that self-identifying members of that category are observed to actually hold.  As I’m going to focus on classifying people according to belief systems, I’ll use as an example the category of Christians.  One sensible-looking prescriptive definition of “Christian” is one who believes in an all-powerful, all-knowing, benevolent deity that created the universe and sent Jesus to us to die for our sins.  Possible descriptive definitions of “Christian” include “person who attends church and donates to the poor”, or “person who reads the Bible, prays, and does their best to love their neighbor as themself”, or “person who wants to curb women’s reproductive rights in the name of the sanctity of life and believes a theistic creation story is more plausible than scientific consensus”.  (Note from this example how widely descriptive definitions can vary according to the points of view of the ones employing them, a point I’ll expand on later.)

The distinction between prescriptivism and descriptivism could be regarded as a choice between on the one hand determining a set of characteristics before finding a class of people which satisfy them, and on the other hand determining a class of people with some apparent commonality before pinpointing a set of characteristics that they share.  Thus, one could view it as a question of which do we fix first: the characteristics or the actual group.

I haven’t seen the question of prescriptivism versus descriptivism applied to categories of people explicitly discussed all that often.  But it appears that for an awful lot of controversies that are being discussed, such differences of opinion or assumptions is bubbling just below the surface.  As the title of this essay suggests, I tend towards the prescriptivist side in most contexts (though not all, as I’ll explain).  But instead of laying out a general justification for this, I think I’ll dive right into a hot-button example.

I. Mere Feminism

Yes, I jumped straight to this topic, because I feel it involves the term whose definition I’ve seen disputed more than any other (with the possible exception of atheist, but that’s a debate with a few more components to it that I don’t want to deal with today).  So let’s get right into it.

The term feminism is incredibly hard to pin down, given that it can refer to a worldview, a social movement, and/or an academic discipline, so instead let’s take a look at feminist.  What do we mean when we refer to someone as a feminist?  Many would respond by alluding to an oft-cited quote and say that a feminist is someone with the radical notion that women are people.  Many others will give an similarly prescriptive definition, saying that a feminist is one who believes that women are equal to men or is an advocate of women’s interests.

But over the last few years, I’ve really been struck with how many people use the term feminist to refer to their perception of the typical self-identified member of the (usually current) sociopolitical movement known as “feminism”.  This is clearly a descriptive definition, not a prescriptive one, and while many of those who adhere to it do so implicitly (seemingly without having considered any other approach), I’ve seen a number of others vehemently defend it against those who espouse the “women are people” one.  Another very relevant thing I’ve observed: almost all the time, these feminist-descriptivists do not themselves identify as feminists; the traits that they claim characterize feminists tend to be negative and reflect what is arguably some of the least flattering feminist rhetoric out there.

This is probably the issue that first got me thinking about the whole prescriptivism-vs.-descriptivism thing a few years back, and for now I’ll just record the main ideas I remember having at the time.  I started out by affirming that clearly we can choose to use words as we like, and our decision has no direct bearing on the true answers to the object-level issues where their use comes up (in this case, actual gender and women’s issues).  However, some choices of word usage lead to more constructive discourse than others.  And in this case, it was (and still is) clear to me that some sensible prescriptive definition is definitely more constructive than most attempts at a descriptive one.  In particular, an unflattering descriptive definition tends in practice to muddy the discourse between those who oppose certain manifestations of modern feminism and those who defend them.

One reason for this is that even those self-identified feminists that believe in some of the things Joe Descriptivist finds objectionable are likely to see Joe Descriptivist’s characterization of feminists as a strawman that doesn’t embody the full range of feminist viewpoints.  A lot of the time, their complaint would be objectively valid, even if we grant that descriptivism is the right approach here.  For instance, as far as I know, the anti-feminist stereotype of bra-burners isn’t based on any group of people that actually existed (though to be fair, I don’t know any anti-feminists today who profess belief in this stereotype).  And oftentimes we are misled into assuming that more extreme views within a movement are mainstream, by the fact that they tend to be propagated by the loudest voices.  More generally, even if all parties accept that feminist-descriptivism is the way to go, it’s inevitable that people will disagree on which sets of views are most representative of the real-world feminist demographic.  And while that question has some significance in the broader discussion surrounding feminism, endlessly arguing over it would seem to distract from issues that are far more central to women’s rights.

Another reason — much more complex and contentious and I don’t want to get myself too distracted elaborating on it — is that for those who object to some particular flavor of feminism they’ve been exposed to, using it to form a descriptive definition of feminist results in them seeing feminism itself as the enemy.  And things tend to go downhill from there — this seems like a recipe for further polarization rather than better understanding.  On the other side, it’s been noted many times over that a good number of self-identified feminists engage in similarly destructive rhetoric by insisting that others who believe in gender equality but aren’t on board with their all of their latest controversial stances can’t possibly be feminists, in a classic example of the motte-and-bailey fallacy (ironically, the author of the linked article, who considers feminist as a main example, appears to be a feminist-descriptivist).

I suggest that a list of beliefs that qualifies someone as a feminist under a sensible descriptive definition might look something like this.

  1. All genders are equal, and in particular women are deserving of equal privileges.
  2. Women’s interests are relevant and should be actively pursued.
  3. Throughout history women have, very broadly speaking as a group, been more oppressed than men, and to some extent this continues to hold.

(I realize that the second statement in (3) is moderately contentious and might not be included in the definitions held by some of my fellow feminist-prescriptivists, and I admit that I hesitantly included it partly to avoid the category becoming too broad.)  Anyone whose belief system satisfies the above criteria plus some other things that don’t necessarily directly follow from them can easily enough be described using a phrase like “intersectional feminist”, “radical feminist”, “second-wave counterculture feminist”, or “millennial feminist pushing for all men to repent for their contribution to toxic masculinity”.

I’ve always connected this debate to C. S. Lewis’ approach to defend Christianity to the public with Mere Christianity, which is perhaps the most celebrated and successful of all time among works of Christian apologia aimed at the layman.  Now it’s been quite a few years since I read that book, but my impression of Lewis’ main strategy as well as an explanation for the title is that he was going out of his way to keep the definition of his religion small (as well as, of course, prescriptive).  That way, instead of getting mired in arguments over which branch of Christianity is the One True Christianity, we can instead focus on debating the core beliefs; meanwhile, Christianity itself can be more appealing and inclusive while the finer points of contention between different Christians are treated as secondary as they should be.  If we want to discuss groups of Christians who hold additional beliefs not directly implied by Lewis’ basic criterion, terms like “Catholic (Christian)” or “episcopalian (Christian)” easily enough get the job done while allowing for more nuanced categorization.

As someone who himself firmly believes in the tenets (1)-(3) listed above, wants to explicitly encourage others to embrace them, and feels that they ultimately deserve a higher degree of importance than the disagreements between different types of women’s rights supporters, I’m all in favor of taking the C. S. Lewis approach when it comes to characterizing feminists.

II. Representation of the Nerds

Now let’s go to another category of people, one for whom the category name is much less overtly contentious but I’ve started to ponder it over the last several years.  I believe it’s safe to say that by pretty much anyone’s definition, membership in this category applies to the type of person likely to be reading this blog post, as well as to its author.  I’m talking, of course, about nerds.

The term nerd itself hasn’t seen a whole lot of debate as far as I can tell, but I have a feeling that those who discuss nerd culture are often assuming subtly but crucially differing definitions.  I myself have lately begun to notice a usage of “nerd” among members of that special subset of the nerd community known as the rationalist-sphere which bothers me.  I think I summed it up pretty well in this post, from which I quote:

Maybe part of what bothers me is the fact that so many seem to treat “nerd”, “introvert”, and “autistic” as interchangeable and all of those things as typically traits of quiet underdog heroes that society doesn’t like. Not that I deny how widespread the ignorance and prejudice is against those who are less neurotypical or less socially competent, but the traits that directly highlight someone’s nerdiness don’t imply those things nearly as much as some make it sound like they do – indeed, IME the definitive nerd traits are celebrated among adults in today’s world at least as much as they are (often affectionately) made fun of.

Not long after I wrote that, a post entitled “The Nerd as the Norm” appeared on the blog Everything Studies.  While its central thesis wasn’t any particular definition of nerd (and I quite like its main point and the way it was presented; I recommend checking the post out), the author John Nerst was explicit about employing a list of criteria for nerdiness that I feel strong reservations about.  To me, these traits mostly read as either variants of “not physical-activity oriented” or mild versions of classic high-functioning autism characteristics (and indeed Nerst placed “autistic” at the extreme end of his nerdiness spectrum).  I honestly surprised myself with the length of the comment I left directly objecting to this.  The gist is contained in the following excerpt.

Your description seems to involve a list of traits that, according to your observation, tend to cluster and are frequently associated with nerdiness. I’m a little uncomfortable with some of the specific choices of traits, because I’m not sure they’re as closely correlated as you seem to think […]. In particular, I’ve been getting frustrated lately at how often people seem to carry an implicit “nerdy = abstract-thing-oriented = introverted (or not people-oriented) = autistic” assumption, when according to my experience, the correlation is not that strong and I don’t particularly care for the stereotype.

His stance (which I hope I’m characterizing fairly with respect to the present context) essentially boiled down to saying that, in his observation, certain personality traits seem to cluster, and that it makes sense to designate nerd to that cluster — in particular, it would be more compatible with the existing social connotations of calling someone a nerd.  My stance essentially boiled down to disagreeing that these traits cluster so strongly around those who would be normally identified as nerds (e.g. among my past social groups that were very overtly nerdy, most members were heavily invested in athletic pursuits and many were quite extroverted and people-oriented), along my suggestion that a usage based on any perceived cluster will perpetuate a stereotype.  I proposed my own definition of a nerd as simply one who is passionate about intellectual pursuits, a definition I’ve assumed for years without question until recently.

Note that as with feminist, I’m insisting on a prescriptive definition while others prefer a descriptive one.  In this case, it’s based on what they view the average real-world nerd to actually be like, often involving introversion, dislike of / disregard for / difficulty with social conventions, being straight-laced, and having a general lack of interest in physical activities (and even, according to one high-profile rationalist nerd, a high degree of scrupulosity).  We see a major contrast from what I emphasized in the feminism example though: here the descriptive definitions I’m disagreeing with are coming from fellow nerds who are even using them in the course of defending the nerd community.  (I’m sure there are plenty of non-nerds out there who use them to attack nerds, but somehow I’m just not as exposed to those voices.)

So why do I, a blatant nerd and certainly not apologetic about it, object to certain common portrayals of nerds by fellow members of my community who are also clearly pro-nerd?

Well, one reason, which should already be apparent from what I’ve just been saying, is that there will always be contradictions between each of our perceptions (usually coming from personal experience) of how people cluster.  What one of my fellow nerds might see as an eloquent portrayal of the purest form of True Nerddom, I might see as an inaccurate stereotype.  But then there’s stereotyping itself, which is where my deeper objection comes from.

The fact is that I’m quite wary of stereotyping of any group, period.  I’m adverse to it regardless of whether or not it comes from a group member who believes that stereotype to be positive, or even if I for the moment feel it’s positive.  Because in my view, most types of social concepts, including stereotypes, don’t carry Goodness or Badness as a fundamental property.  To quote myself from the essay I just linked to:

[C]onsider the classic stereotype of Asians being good at math.  That may seem harmless for Asian people at first glance. But what happens when people start taking it seriously enough that non-Asians automatically expect the Asian person they just met to excel at math and feel inclined to tease them for being a “bad Asian” on finding out they don’t, or if an Asian person feels bad for not being especially good at math, or if non-Asians start routinely expecting Asians to help them with math problems, or if negative stereotypes that are already attached to “math people” become associated with Asians?  The fact is that all flavors of stereotyping are ultimately harmful in that they involve over-generalizations among certain categories of people and lead to invalid assumptions that are guaranteed to create pain in the long run.

Don’t get me wrong.  I have nothing against introverts or black sheep who, in the social arena, play to the beat of their own drums.  I have no problem with autistic people (while certainly having a big problem with ableist attitudes directed against them).  Those who are romantically and sexually unsuccessful and awkward around acquaintances they’re attracted to are generally fine people in my book and certainly an improvement over confident creepers who carry out sexual conquests through unscrupulous manipulation.  And I consider being a meek and mild, well-behaved homebody who prefers to have a few friends over to play board games on Friday nights to be an equally valid (if not overall healthier) lifestyle to dedicating one’s weekends to clubbing or attending tailgate parties and football games.  A lot of nerds do have these traits, and they deserve defending against those who sneer at them for it.

But on the other hand, I can’t bring myself to classify the above traits as absolutely, positively Positive.  There are ways in which I’m far more socially conventional than the stereotypical nerd, and I don’t appreciate being treated with the assumption of social strangeness (and anyway there’s a fine line between perceptions of “strangeness” and awkwardness or incompetence).  The autism spectrum lies under a wide umbrella, but the public perception of high-functioning autism is often skewed (perhaps by fictional novels like The Curious Incident of the Dog in the Night-Time) towards some rather unsavory edge cases of autists with sociopathic tendencies.  I wouldn’t want such anti-autistic stereotypes associated with me, plus there’s that whole social awkwardness/incompetence thing again.  There are also situations where I find it really less than helpful to be perceived as asexual (possibly via a mental pathway from “nerd” to “Aspie” to Sheldon Cooper).  And moreover, there are times when I want to be treated as a chill guy who’s down for going out late to a bar, grabbing some beer, and mingling with the crowd with less-than-high-brow conversation, as opposed to someone who’s only able to socialize with game controllers, dice, or Magic the Gathering cards in hand.

There are nerd stereotypes out there which are presented in positive and negative lights, and I fit some criteria of each pretty well while not fitting some others.  But given the choice between a definition for the term itself that involves any of them, versus a more basic criterion that seems to capture the fundamental essence of nerdiness in a way that fits most people’s intuitions without considering possible correlated traits, I’m going with the latter.  If we do want to specify some additional correlated trait, phrases like “introvert nerd”, “autistic nerd”, “gamer nerd”, “straightedge nerd”, etc. will do fine.

III. A closer description of my prescriptivism

Looking at the above two examples, my reasons for preferring prescriptive labels over descriptive ones can be summarized as follows:

  1. Descriptive definitions tend to rely on the personal experiences and impressions of the ones who come up with them and can be very hard to agree upon; settling on a descriptive definition of some group tends to require a lot of arguing over various trends among self-identified members, which can be a distraction from more relevant issues.
  2. Descriptive definitions for groups perpetuate stereotypes about members of those groups, which are always dangerous regardless of whether they’re used by outsiders to attack it or from within the group to defend it (or to make their group more exclusionary).
  3. By agreeing on a reasonable prescriptive definition for a category, that category generally becomes somewhat more inclusive while still allowing for concise terminology for each of the subcategories people want to talk about, and this makes for clearer and less polarizing dialog.

Before I wrap up, I want to make a couple of major caveats.

First of all, I meant it when I said right at the top of this post that I have a fairly low level of confidence in my position here.  To clarify, I actually feel pretty confident that the points I’ve made do carry a lot of validity, but I’m not sure I’ve considered the whole story.  I can imagine some potential pitfalls in the prescriptive route also.  For instance, I’ve established that it can be hard to agree on an accurate description of what most self-titled members of a certain category are like, but I’ve given no formula for agreeing on a prescription for their category that goes beyond “choose some fundamental characteristic lying in the overlap of our intuitions”.

I believe that for feminist and nerd, such fundamental characteristics do exist, but even there I have to admit some room for dispute, such as in the last part of my proposed list of criteria for qualifying as a feminist.  There are other categories for which I’m less sure of this.  For instance, how does one come up with a definition for hippy that isn’t at least somewhat descriptive (leaving aside the fact that original group members seem to consider the term mildly derogatory)?  My first instinct is to characterize the hippy subculture in terms of a socially liberal, anti-materialist lifestyle that involves incorporating nature into as many facets of life as possible and reflecting passion for environmental causes, animal rights, and pacifism.  I’m not sure how far along the prescriptive-descriptive axis that definition is.  I was trying not to be too descriptive and stick to the deeper-rooted ethical principles (note that I left out long hair, recreational drug use, living on communes, etc.), but already at least one trait I mentioned (regarding animal rights) doesn’t hold up that consistently with the first-hand stories I’ve heard from members of the original counterculture movement.  Do I include in my list of criteria as many “often”s and “usually”s as possible, or does that make everything too wishy-washy?  Do I accept that when dealing with subcultures as opposed to social movements I have to be somewhat descriptive (except that nerd clearly refers to a subculture and many components of hippy-ism are obviously part of a social movement)?  I’m leaning towards a rule of thumb where I try to be as prescriptive as possible but recognize that it’s easier in some cases than in others.

Moreover, it’s important to take in account that semantics change across the decades, which is unfortunately trickier for the prescriptivist to deal with than for the descriptivist.  The word nerd originates in 1950 from Dr. Seuss’ If I Ran the Zoo, where it refers to one of his quirky anthropomorphic fictional creatures.  It apparently acquired a meaning of “bookish person” pretty quickly.  However, a memory recently came back to me of seeing and reading Larry Shue’s play The Nerd from the early 1980’s, where the title character is an obnoxious, socially oblivious loser showing no brightness or intellectual passion whatsoever whose job is “making sure there’s chalk in the crates” (warning: this production comes much closer to a depiction of a mentally retarded person than of anything resembling what a modern audience would call a nerd).  My “passionate about intellectual pursuits” definition might only ring true in recent times.

This sort of brings me to my second caveat, which is that there’s one type of category that I consider an exception and for which I’m rather staunchly descriptivist.  That is any label that implies allegiance to an organization.  This is because an organization (I’m thinking in particular of political organizations here) already has a well-defined platform, and the fact that this platform may change from year to year means that words referring to that organization should change meaning from year to year.

Thus, while the term communist should be defined prescriptively, a capital-C Communist understood to mean a supporter of (say) the Soviet Communist Party of the 1950’s should satisfy a descriptive list of criteria.  Or to take another example, if I were to say that I’m a Republican now, obviously referring to the American political party rather than favoring republicanism or something, I would mean that I believe in most of the principles laid out in the current official Republican platform as well as those espoused by current Republican leaders.  I would not mean that I stand for a stronger central government and against slavery spreading to new US territories as the Republican party of 1860 did, nor that I’m in favor of isolationism and a return to normalcy as the Republican party of 1920 did, nor that I cling to some abstract ideal of what a Republican party platform should look like.  I believe this convention is important in making people feel free to leave political parties as their platforms change or as their leaders betray said platforms, and that in fact, stubborn refusal to abandon one’s own unchanging notion of one’s political party is what leads to a lot of unfortunate party tribalism.

Well, that’s more or less where I stand on this matter, which, pedantic as it may seem, does seem to play a role in many conversations on more substantive topics.  I’ll be curious to know what people think of where I come out on this and what further nuances come to my attention.

7 thoughts on “In defense of prescriptive labeling

    1. That is strange. I don’t have an email informing me of your long comment. I vaguely remember that a couple of times my attempts to leave very long comments on other people’s WordPress posts have failed on the first attempt, but I don’t remember exactly how.


      1. To tell the truth, I’m not entirely sure which of my spam folders it would land in, but at the moment none of them contain any message more recent than March. :/

        If you feel up to re-writing your comment, that would be great, but if not I understand — effortful comments never feel like they come out worded as well the second time.


      2. Trying a few separate ones.

        It’s a lot to process. I sympathize, I really do, but I think you’re fighting a losing battle against language if you insist on meanings different from the ones people have in mind when they use words. The family resemblance model seems obviously correct to me wrt the meaning of words, and that means that meaning and definitions is going to be context dependent and organized in a cluster structure, successively expanded in different directions through semi-metaporical use.

        This is of course the case with “nerd”, which has expanded in meaning quite a lot in recent years. Fifteen years ago (ish) when my girlfriend and I had just met she once called me a nerd and I objected. It was an insult, in my mind, referring to a meek, annoying, bespectacled 80’s style poindexter getting his head dipped into the toilet at lunch break. SInce then it’s come to refer to a much larger set of people, often just meaning that someone has interests they care about. And I’m comfortable with it now, even though I, like you, probably would pick sports + beer or even clubbing over MTG (or similar, I can’t play MTG) on most nights.


      3. Cont’d.

        My article used a particular, perhaps old-fashioned meaning of the word corresponding to the article I was making fun of, and I don’t think such local use of words is wrong. I do think there’s something there: that people-oriented vs. thing-oriented is a real and important dimension, that abstract/intellectual vs. concrete/physical interests is one too (maybe you disagree, I admit it’s less obvious), and that the intersection of them kinda-sorta points to the “classic nerd” personality type.

        In general, what you’re grasping at seems some combination of intensional vs. extensional definitions, and essential vs. incidental characteristics – but none of them captures the disinction perfectly. Ultimately it seems to be a question of wether you should form categories by reifying the lumps in thingspace, or by carving it up by decree. Both have their uses and the issue won’t be solved any time soon 🙂

        Liked by 1 person

      4. I mostly agree, actually, in particular that I seem to be fighting a losing battle, and also that the people-oriented vs. thing-oriented is an important distinction.

        As you suggest, this is obviously a question of how we should carve thingspace into categories. It’s therefore, compared to most debates, less of a disagreement over objective truth and more of a disagreement over what is the best possible use of language from a pragmatic standpoint, with the purpose of enabling clear thinking and valid judgments. (Although I’d argue that to some extent, many if not most debates — e.g. free will vs. determinism — are debates over what terminology is more pragmatic, but prescriptivism vs. descriptivism is more transparently so than most others.) Therefore, actually defending one side or the other amounts to arguing that humans will process and harness language in certain ways, which is a rather imprecise business involving insight into a lot of subtle psychology. My post was less an attempt to demonstrate the objective validity of the prescripivism side, and more a personal expression of “This is a dilemma we should all be aware of and decide which side we stick with when; here’s the side I stand by most of the time (somewhat for subjective emotional reasons)*; let me list some of the advantages of that side.”

        * I do feel my stance on “feminist” comes from a more objective place (I feel almost alone in taking it but think there are good points there that more folks should consider regardless of their views on gender relations) and arises less from personal feelings than the part about “nerd”, though.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s