All my marbles in one place

A blog about course design, data display, C++, and Python. Yes, these are related, at least as I see them.

© 2021. All rights reserved.

The 'Descriptive' versus 'Prescriptive' distinction is irrelevant for actual writing

13 Jun 2021 Tags: Rhetoric

I sometimes get asked whether I’m a “prescriptivist” or “descriptivist”, or—worse still—labelled as one. This has always struck me as beside the point. In none of my roles, not as a writer, teacher of writing, or editor, has this distinction ever provided useful guidance. Instead, I base my usage choices upon some combination of assessed ambiguity to the reader, existing convention, degree of formality, emotional impact, and other considerations. The resulting choices might sometimes conform to prescriptive rules or existing usage or neither. Effective writing is a balance of competing requirements, not rote adherence to an overarching “philosophy”.

The distinction, briefly

The Wikipedia page on linguistic prescription summarizes the goals and paradoxes of the prescriptive approach well. It emphasizes the essential point that prescription, the enumeration of fixed rules for “proper” use, and description, the summary of empirical surveys of how language is actually used, are complementary rather than competing. This viewpoint is prevalent in academic discussions of the topic.

Yet outside the academy, “prescriptive” and “descriptive” are often misrepresented as competing rationales for a given writing choice. Does this choice reflect adherence to some standard of “correct” language? The writer made a prescriptive choice. Did the writer choose to deviate, in small or large part, from some standard? The writer made a descriptive choice. If a writer seems to make many choices that might be characterized by one of these terms, then critics infer that the writer adheres to one or the other “philosophies”.

This is ultimately just a play of shadow puppets, where critics never see the writer’s actual process, only some shadows on a wall, and rationalize their explanation by an appeal to some hidden figures supposedly casting those shadows.

When it comes to my own writing however, I do have access to the process underlying my choices and I can say that neither of these “philosophies” (again, which is not how scholars of linguistics apply these concepts) guides my choices. Instead, I’m guided by a much broader range of concerns. I want to illustrate this with two examples, one of which might appear “prescriptive”, the other “descriptive”.

Example 1: The possessive case of “it”

I consistently (well, as consistently as I can muster in the face of autocorrection mistakes that I don’t catch) write the contraction of “it is” as “it’s” and the possessive case of “it” as simply “its”:

In a strictly surface sense, this is a prescriptive choice. Any stylebook or manual of English usage will recommend this and it is sanctified by longstanding and widespread practice in edited writing. Yet my reasoning behind this choice is not simple rule-adherence, though the existence of the rule informs my choice.

The problem

The problem arises from the collision of two widely-used punctuation conventions of edited English:

  1. Form the contraction of a noun and “is” by the “'s” suffix.
  2. Form the possessive case of a noun by the “'s” suffix.

We apply both rules routinely for nouns, with no apparent ambiguity:

For most pronouns, instead of a “'s” suffix, the possessive case is denoted by a different form:

Only the impersonal first-person pronoun “it” never acquired a distinct possessive form. Explaining why this happened, if there is any explanation more sophisticated than “English be weird”, is beyond my skill in linguistics. For our purposes, the main point is that no such form became widely accepted and so a convention was required.

Editors could have standardized on the same convention used for nouns, namely use the “'s” suffix for both:

Instead, the overwhelming majority of editors chose to distinguish the two possibilities by restricting the “'s” suffix to indicate the contraction and simply using an “s” suffix to indicate possession:

That’s the choice the editors have made. Why do I choose this as well?

First, let’s distinguish writing for an edited publication from writing for self-publishing (as in this blog). For an edited publication, the principle is simple: Follow whatever house style their copyeditors require. Their house, their rules. Simple enough.

But what about writing for self-publishing on the Internet or for internal organizational documents, where I get to set my own standards?

Principles favouring this convention

The following principles support the convention:

  1. Adhere to widely-used conventions.
  2. Preserve distinctions.
  3. Reduce cognitive load for the reader.
  4. Support fluent readers.
  5. Avoid accusations of solecism.

Consider each in turn.

1. Adhere to widely-used conventions

All other things being equal, work within convention rather than flout it. There are appropriate times and places for working against convention but the outcome must warrant it. I don’t see any substantive gain from apostrophizing the possessive “it” for the sort of writing I do. This isn’t the inner monologue of a distracted fictional character.

2. Preserve distinctions

Once again, all other things being equal, preserve distinctions rather than effacing them. A contraction is distinct from posession, grammatically and semantically, so it is useful to represent them differently in writing. This logic in fact argues against the convention for nouns, which does not distinguish the cases, but in that case, the first principle, adhere to common convention, applies.

3. Reduce cognitive load for the reader

This typographical distinction is more valuable for the impersonal pronoun than for nouns because a pronoun imposes more mental effort, requiring the reader to connect it to its referent, whereas a noun is self-contained. The phrase “Norm’s performance” is more clearly a possessive form because it names the possessor, whereas “its performance” requires the reader to recall the referent of “it” to determine the item whose performance is under discussion. Compounding this cognitive load with the requirement to disambiguate the possessive case from a contracted verb might burden the reader. Eliminating the ambiguity makes the load more manageable.

4. Support fluent readers

Studies of fluent readers show that they read entire phrases rather than individual words and that they often predict the completion of phrases before they have reached their end. The widespread distinction between contraction and possession of “it” in turn leads to fluent readers expecting it. A sentence that does not conform to this convention will disrupt the flow of a fluent reader whereas a sentence that does conform will support that flow.

5. Avoid accusations of solecism

Usage purists who call out every small deviation from received usage, such as

and consider such usage to be solecisms. There’s a subtle trick in this choice of term that is worth expanding in a bit of detail.

Webster’s offers three related meanings for “solecism”, ranging from “ungrammatical” to “a breach of etiquette” to “deviating from the proper, normal, or accepted order”. The Oxford English Dictionary [subscription required] offers three very similar meanings. Many critics of writing or critics of a given article exploit these shades of meaning, conflating “nonadherence to a convention” with “rudeness or improper behaviour”. Note that writing “it’s performance is good” isn’t even ungrammatical. It’s simply a nonstandard representation for the possessive case, a different way of putting the underlying grammar into letters. Calling it a solecism is a cheap rhetorical trick, misrepresenting a nonstandard usage as an indicator of the writer’s lack of intelligence, education, effort, or probity.

I don’t agree with the characterization of “it’s performance is good” as a solecism but I acknowledge that criticism’s existence. Once again, all other things being equal, best to head off such cheap critiques in advance. There are times and ways to strategically piss off grammatical prigs but this is rarely one of them.

Combining these principles

In this case, multiple principles that I use to guide my writing agree on a single outcome, the one that “prescriptive” purists require. Though my choice is not based upon prescriptivism, the choice is influenced by it. The near-universal use of the prescribed convention in edited English creates a context where there are genuine gains from conforming and none from violating it.

In a moment, I’ll provide an example of a choice where the gains from violating the prescribed choice outweigh the risks, but first, as long as I’m on the topic, I want to toss a few stones at an egregious example of prescriptivism gone wrong. Yes, it’s time to discuss The Elements of Style.

The obligatory critique of The Elements of Style

One of the most influential prescriptive texts in the US context is Strunk and White’s The Elements of Style. Much of the book lists Rules That Must Be Followed, with no supporting rationale. The absurdity of this endeavour is unintentionally revealed by White in his introduction:

Some years ago, when the heir to the throne of England was a child, I noticed a headline in the Times about Bonnie Prince Charlie: “CHARLES’ TONSILS OUT.” Immediately Rule 1 lept to mind [which requires “CHARLES’S TONSILS”]… Clearly, Will Strunk had forseen, as far back as 1918, the dangerous tonsillectomy of a prince (3rd Ed., p. xv]

When I first read this passage, I puzzled over the offending headline for some time, searching for a problem. It’s not ambiguous, it’s not unclear. Every single reader would gather the meaning without effort yet White sees no need to justify his criticism. It’s enough that the headline violated Rule 1, laid down for all time by Will Strunk in 1918.

I note that at least by 1966 the Times stylebook no longer accepted this heinous solecism:

That Prince Charles’s [sic] royal wages apparently have doubled as a consequence of his 18th birthday did not pass unnoticed among the country’s politicians (Times, 15 Nov 1966, p. 16)

Writing instructors have become so tired of the imperious rule-binding tone of Strunk and White that some authors have taken to satirizing it in the titles of their own style books:

There are useful lessons from The Elements of Style but I would never recommend it as anyone’s first book on style. For my money the best starting point, a book guiding my own writing daily, is Joseph William’s Style: Toward Clarity and Grace. Read it through at least twice and practice the recommended techniques before moving on to Strunk and White. And if you never do move on, it’s not a big loss.

And now to the next example, where I happily break a “rule”.

Example 2: A deliberate “comma splice”

The first example was a case where my choice matches the recommendation of a staunch prescriptivist (again, as typically misconstrued by nonlinguists). In this second example, I present a choice that violates prescriptivist rules.

In the right context, I am perfectly fine writing comma splices, a particular bête noire for prescriptivists. Consider the case where I am describing a debugging session. As part of the testing, I started a service and noted that it didn’t run. I could combine these two steps in at least four ways:

  1. I started the service. It didn’t run.

  2. I started the service; it didn’t run.

  3. I started the service and it didn’t run.

  4. I started the service, it didn’t run.

The first form is simply two short sentences. The next two forms, a semicolon and a conjunction, are widely accepted forms within formal edited English. The last form is—horrors!—a comma splice. Why would I choose it?

1. Enact an experience

To my eye and ear, each of these four reads differently and has a different effect. The two-sentence and semicolon forms are stoccato, with a slightly smaller pause at the semicolon than the period. They emphasize the discrete nature of the statements: An action. An observation.

The third form, joining the statements with the conjunction “and”, presents them as a single unit, emphasizing the result (it didn’t run) and presenting the action (starting the service) as a precursor.

The last form, joining the statements with a comma, enacts an experience. The form is a bit breathless, suggesting that I was tired, that this debugging session had gone far longer than I expected and I just wanted to get it done.

Beyond the expressive aspect, there is another justification for this choice: It is in fact “grammatical” to elide conjunctions. The technique is called asyndeton.

2. Use asyndeton for impact

Eliding conjunctions to increase the forcefulness of a statement is such a well-established technique that the standard name for it dates back to the ancient Greeks. It makes a statement seem more direct, even blunt:

Both forms are entirely acceptable in formal writing but the second is more emphatic, confident. To be effective, the technique needs to be used sparingly. When used frequently it loses force and can be tiring in the way that repeated deviation from convention breaks the reader’s flow and draws attention to itself. The first principle in Example 1, “Adhere to widely-used convention” applies here as well. But variation against that conventional backdrop sustains interest.

Combining these principles

In this case, two completely different principles guide my choice, overriding several of the guiding principles of Example 1. The result violates precriptivist dicta and risks charges of solecism but I judge the effect worth that risk.

Simply select the usage that best supports your argument

Writing is ultimately about making the most persuasive case to a specific audience. When making decisions about usage, leave philosophies of language to the subset of linguistic scholars who debate such things (and have a more subtle view of the issues than this in any case). As a writer, my concern is what will make my point most directly and produce the strongest effect. Other considerations simply distract from that fundamental task.