Following up on my recent “zombie rule” postings (here and here): there are two sorts of phenomena here, though I’ll argue, eventually, that the difference between them isn’t important for practical purposes.
I’ll start with the history of a typical zombie rule:
Stage A: In some context, only variant X occurs. People do not use variant Y in this context, so they are in some sense obeying a “rule against Y” — though this is a deeply misleading way to look at things, since these people don’t really have a “rule proscribing Y” (an implicit rule), they just have no (implicit) rule allowing Y.
Stage B: Variant Y is innovated in the relevant context. Now some people have Y in this context, so for them there is no longer a “rule proscribing Y”.
Stage C: Usage critics attempt to restore the status ante quo by explicitly banning Y (in the relevant context) — a tactic that very rarely succeeds.
Stage D: Usually, the innovative variant Y spreads, and the explicit proscription of Y becomes a zombie rule. [But sometimes Y recedes, as a short-lived fashion (quotative be all seems to be receding, fairly dramatically, in favor of the — equally non-standard, or at least colloquial — quotative be like), though not because speakers and writers are listening to the usage critics.]
Such zombie rules can have a long life-in-death, often supported by specialized advice literature — for children, teenagers, or college students; for business writers; for journalists; and so on. Not infrequently, people who’ve been exposed to such advice cling to zombie rules passionately, probably because the proscriptions seem arbitrary and unnatural to them, and consequently demand considerable work to learn and constant vigilance to maintain; zombie rules are hard-won “knowledge” and are precious for that reason alone. (A number of people have suggested that such considerations lie behind the passion many have invested in correct spelling in English.)
Here’s another, somewhat different, scenario, in which we jump right in at an analogue of stage C above. The background involves a long-standing variant Y (in the standard language) that someone takes a dislike to, on the basis of some “theoretical” consideration, or simply as a matter of taste. So a “rule” proscribing Y is promulgated; this is a proposal to change the standard language, by eliminating some existing modes of expression.
The Language Loggers have posted at some length about several examples of this type: the “rule” banning stranded prepositions, the “rule” banning sentences beginning with coordinators (in particular, and and but), the “rule” against however as a sentence-initial discourse connective, and so on.
Both sorts of rules are proscriptive, but the first type proscribes variants that are perceived to be innovative (and therefore non-standard, or at least unacceptably colloquial), while the second type proscribes long-standing variants. The first is reactive, the second proactive. For the first, there was a time when the now-proscribed variant simply wasn’t around (so that no one articulated a rule against it), while for the second, the now-proscribed variant has a considerable history.
[As with the classic zombie rules, once the second sort of rule is out there, it’s hard-won “knowledge”, taking real effort of learn and real effort to monitor, just because it’s not “first nature” for speakers and writers — though, of course, such things can be internalized, can move (with enough practice) from consciously monitored production to “second nature”.]
Some time ago, Geoff Pullum pointed out to me that it’s a bit odd to use the zombie metaphor for the second type of rule, since the rules were never alive. Well, not until someone breathed life into them; they are, in effect, Frankenstein’s monsters that became zombies.
In a February 2006 talk at the Radcliffe Institute for Advanced Study (at Harvard), Geoff drew the distinction between the two types and invented terms for them: (true) zombies and bogeymen, respectively — what I think of as “natural zombies” and “constructed [or concocted] zombies”. What Geoff said in that talk was:
The difference between a zombie and a bogeyman is that while zombies used to be alive but are now dead, bogeymen never existed at all…; they are just mythical beings invented to frighten children.
Later that year (in November), Geoff and I had an exchange on this point. Here’s my side (slightly edited):
I was looking for a general term for rules that have no validity, no real life, regardless of their origin. Some once had a real life but died, others were granted some sort of twilight existence by their creators, who of course intended to bring a fully living being into the world but succeeded only in creating a monster. From the point of view of current users of the language, they’re the same sort of creature. So I want a general term that takes in both of them.
But I also want to be able to make the distinction, and I don’t have good terms for distinguishing natural zombies (your zombies) from constructed zombies (your bogeymen).
By the way, I’ve never had much success with the tactic of explaining to people that some zombie rules are entirely constructed, that they never actually governed the way (the best) people spoke and wrote. I get two responses: (a) regardless of the history, the rule is a good one (and here my interlocutor cites virtues of the rule); (b) regardless of the history, the rule now governs the way many, or at least some, people speak and write (that is, the rule has become part of the ideology of the language to the extent that it actually affects the way people behave). I’m prepared to counter response (a) on the facts, though there are sticky points.
Response (b) is trickier to deal with, because in many situations the facts of history are not particularly relevant to the way social life is configured. Opinions, attitudes, and (folk) beliefs are facts too, social facts. Bill Bright’s discussion of squaw and its subjective associations makes this point very clearly; by now, its actual etymology should probably be beside the point for a thoughtful and well-meaning person.
Along the same lines, many African Americans are absolutely sure the positive term phat (so spelled) has an acronymic origin, though people differ as to exactly what that was — I have collected about a dozen candidates, all of them sharing reference to physical characteristics of women’s bodies (the term then being extended to many other things). Some of these people, including highly educated people, recall having had this etymology explained to them in childhood, sometimes more than fifty years ago. For them the acronymic derivation is fact; it’s part of what the word conveys to them.
A different sort of example: famous quotations are very frequently re-shaped to make them pithier or to make them conform to modern usage. But only a egregious pedant goes around insisting on “glister” rather than “glitter” in the gold quote, or on “once more unto the breach” rather than “once more into the breach”, etc.
Still another sort: particular usages are often divorced from their historical sources by the processes of historical change. For instance, the use of (non-standard) all’s in “All’s I know is that…” is, historically, a contraction of all as, with a (non-standard) complementizer as rather than that (as in “I don’t know as I’d do that”). It’s not hard to collect all’s and complementizer as in modern English. But there are plenty of people who use all’s but do not use complementizer as, and if you tell them that all’s is really (for some value of “really”) all as, they flat-out reject the idea; for them, there is no as in all’s, which is simply a word on its own (very often spelled alls, tellingly).
It’s hard to tell when the facts of history should be germane to a question of usage.
My main point in all of this is that Geoff Pullum’s distinction between zombies and bogeymen, fascinating though it is to linguists, is probably beside the point for ordinary people. What ordinary people need to know about is the distinction between valid and invalid advice (and zombies and bogeymen are both species of invalid advice).