In a 1968 paper by Kiparsky ("Linguistic universals and linguistic change"), a historical-change argument is made for the brace notation of SPE, based on the history of vowel shortening. The premise is that there was a single rule in Old English which shortened a long vowel before three consonants (gōdspell → godspell; brǣmblas → bræmblas), and also shortened a vowel which is followed by only two consonants as long as it is in the third syllable from the end of the word (blēdsian → bledsian). It is claimed that there is one rule in OE which does this: V → [-long]/ __CC{C,...V...V} with braces combining "a third consonant, or two vowels".

It is claimed that Modern English has a related rule, V → [-long]/ __C{C,...V...V}, shortening a long vowel before two consonants or a consonant followed by two syllables, and moreover this rule arose in Early Modern English. An alternative view is that there are simply two separate rules. Kiparsky uses this rule change as evidence for brace notation, on the grounds that it would be a "peculiar fact that two separate, unrelated rules have undergone an identical modification at the same point in the history of English".

Two factual claims are made: there was a uniform change in the nature of the two rules, where for both subcases, "two consonants plus" becomes "one consonant plus"; this happened at the same time. The subsequent literature has, to the best of my knowledge, not critically re-investigated this claim (there are sporadic counterclaims that the sub-changes were at different times, with no supporting data). I am aware of the competing analysis that the change has to do with a change in syllable structure – the deeper question is whether the underlying factual claim is actually well-supported?

Has any evidence at all emerged since then that either supports or refutes Kiparsky's claim of "same change at the same time"?

  • 1
    I don't see why it matters whether the two subcases came into effect at the same time. Each new generation of children has to infer what the rules are from what they hear, because of course there are no wise teachers to explain SPE notation to them. Perhaps after coming to lax vowels before CC there was a period of confusion and fluctuation before some exceptionally bright children realized they had to do it with braces and invented trisyllabic laxing as we know it today.
    – Greg Lee
    Commented Jan 21, 2018 at 22:12
  • If they came into existence at different times, that is evidence that the two subcases are not a single generalization, as posited in SPE. The claim is that the rule with brace notation was already present before any changes into Middle English. K's argument putatively introduces external evidence for the SPE notation, and is itself used as justification for the notation.
    – user6726
    Commented Jan 21, 2018 at 22:34
  • 3
    But you're treating historical changes as though they were changes operating on rules. This is not plausible. Even if you're a generative phonologist, you can't think that. Children can't see rules -- they have to infer them from the forms they hear.
    – Greg Lee
    Commented Jan 21, 2018 at 22:53
  • 1
    No, I am not: Kiparsky is. Anyhow, the question is not about whether braces are empirically motivated, it is an extremely specific question of fact, about the known history of English phonology.
    – user6726
    Commented Jan 22, 2018 at 0:03
  • "there are sporadic counterclaims that the sub-changes were at different times, with no supporting data" - more details please?
    – Alex B.
    Commented Jan 22, 2018 at 4:17


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.