Roberts/HT 2014 Week 6


Theories of Phonological Change

Ohala (1989; 1992; 1993): Hypocorrection and Hypercorrection.

There are two logical possibilities for ways this can go wrong, and Ohala claims both of these are attested in language change:

Hypocorrection — Bob mistakes noise for signal.

Example: Allophones of /ɡ/ in Catalan

Recall from our exercises in week 3 that /ɡ/ has 3 allophones in Catalan:

  • [ɡ] in onsets (e.g. [ɡat] ‘cat’).
  • [ɣ] between vowels (e.g. [ɡrɔ.ɣə] ‘yellow-f’).

How did this come to be? Well, there are articulatory pressures towards it: maintaining a full stop closure at the velum is comparatively difficult, because the space between the closure and the larynx is relatively small, so supraglottal pressure rapidly forces the closure open again.

Ohala’s story here would be that at some point in the history of Romance, it was common for your average Catalan Alícia to produce an intervocalic /ɡ/ with a relatively early relaxation of the velar closure, so that a large number of Robertos (Catalan Bobs) misinterpreted this as an intentional token of [ɣ].

Hypercorrection — Bob mistakes signal for noise.

Example: more Catalan

Catalan has final devoicing, so [k] can also be an allophone of /ɡ/ in codas [ɡrɔk] ‘yellow.m’.

Given the relative lack of prominence of the coda position, Roberto may either fail to perceive any of the voicing of a final /g/ that Alícia intended to put there, or he may discard a perceived voicing cue as too weak to be sure of.

Further example: dissimilation of labialized velars in Romance

E.g. Latin kʷinkʷe > Italian /ʧ/inque ‘five’

At some point in history, listeners mistook the labialization of the first top as arising from anticipation of the labialization of the second.

Blevins (2004; 2006): change, chance and choice

change — The phonetic signal is misperceived by the listener due to acoustic similarities between the utterance and the perceived utterance; and biases of human perceptual system.

Alice says [anpa], Bob hears [ampa]

chance — The phonetic signal is accurately perceived by the listener but is intrinsically phonologically ambiguous. The listener associates a phonological form with the utterance which differs from the phonological form in the speaker’s grammar.

Alice says [ʔa̰ʔ] for /aʔ/, Bob hears [ʔa̰ʔ] for /ʔa/

choice — Multiple phonetic variants of a single phonological form are accurately perceived by the listener. The listener (a) acquires a proto-type or best exemplar which differs from that of the speaker; and/or (b) associates a phonological form with the set of variants which differs from the phonological form in the speaker’s grammar.

Alice says [tuʔəlaŋ], [tuʔəlaŋ], [tuʔlaŋ] for /tuʔəlaŋ/, Bob hears [tuʔəlaŋ], [tuʔəlaŋ], [tuʔlaŋ], and assumes /tuʔlaŋ/

Kiparsky (1965; 1968; 1973; 1988; 1995; 2008; etc., etc.): New Rules

The paradigm case is easy: these days, after all, we often notate Neogrammarian sound changes with rewrite rules, as for :

To model a sound change in generative phonology, surely you just change a > to a →, and you’re done?

Well, there’s one open question still: where in the derivation do you put the new rule?

Kiparsky (1965) argued that generally the rule is added so that it applies last, but it is possible to add it elsewhere, e.g. in Lachmann’s Law, a rule in Latin...

So we get lengthening in e.g. āctus ‘done’ (< PIE *h2eg-) and lēctus ‘chosen’ (< PIE *leg-)

But not in iŭssus ‘ordered’ (< PIE *i̯udʰ), or sĕssum ‘sit-supine’.

Kiparsky’s solution: the actual conditioning environment is the voicing of the underlying consonant (which is recoverable from the rest of the verbal paradigm, e.g. in agoāctus). There arises a new lengthening rule:

V → [+long] / +obstruent +obstruent
+voice voice

Crucially, this rule is not added to the end of the derivation, but in a counterbleeding relation with the voicing assimilation that gives us a voiceless C in ā[k]tus

[+obs] → [α voice] / +obstruent

This analysis originally appeared in Kiparsky’s doctoral thesis (1965). It is repeated in King (1969:43ff.). For a more up-to-date summary of the literature, see Jasanoff (2004), Roberts (2009), or even Roberts (2012: ch. 3).

Only one other putative instance of rule insertion has ever been proposed, by Watkins (1970), and also from Latin. No others were forthcoming, so that eventually King (1973) decided that rule insertion was not possible after all!

Bermúdez-Otero (1999; 2003; 2006; 2007; 2013): Amphichronics (see also Kiparsky (again!) 2006)

To arrive at a full explanation of a phonological phenomenon, it is often necessary to know something about both its present facts and its history.

At the same time, we have to remember that children do not do comparative reconstruction in their heads when acquiring language: nightingale is almost certainly not underlyingly /nixtiŋɡaːl/ (pace Chomsky and Halle 1968).

Amphichronic phonology is centred around theories that incorporate the life cycle of phonological generalisations into their architecture, such as Stratal OT and Lexical Phonology:

Sound change begins when a phonetic tendency is codified as a phrase level rule. Sandhi processes make then environment of the rule more stable word-internally, so the rule is acquired at the word level. Eventually, morphological alternations force the rule into the stem level, and finally it may cease to be productive at all, so that its residue is visible as a systematic property of lexical entries.

The diachronic effect is domain narrowing: the domain of a phonological generalisation can be observed to shrink over time. This is recorded e.g. by James Elphinston (18th C.) for post-nasal /ɡ/-deletion in English (see Bermúdez-Otero 2011: 2024):

StageRealization of underlying /ŋg/Level reached by the ruleVariety of English
elongateprolong-erprolong itprolong
0ŋɡŋɡŋɡŋɡEarly Modern English
1ŋɡŋɡŋɡŋphrase levelElphinston (formal)
2ŋɡŋɡŋŋword levelElphinston (casual)
3ŋɡŋŋŋstem levelPresent-day English


We are now in a much-debated area of the wild academic frontier: practical exercises are hard to set since there is no consensus on what the answers would be.

Therefore, in lieu of exercises this week, I suggest that you read through at least those papers from the reference list that are marked as optimal. Please spend your class discussing the state of debate in the field. You may use the following questions as jumping-off points:

I leave the division of these questions among yourselves, and the general form of the discussion, to you, but remember: I want a clean fight. Nothing below the belt, and no eye-gouging. Holly will referee.


Blevins, Juliette (2004) Evolutionary phonology: the emergence of sound patterns. Cambridge University Press.

Blevins, Juliette (2006) “A theoretical synopsis of Evolutionary PhonologyTheoretical Linguistics 32:117‒116.

Bermúdez-Otero, Ricardo (1999) Constraint interaction in language change: quantity in English and Germanic. PhD thesis, University of Manchester and Universidad de Santiago de Compostela.

Bermúdez-Otero, Ricardo (2003) “The acquisition of phonological opacityin Spenader, Jennifer; Anders Eriksson and Østen Dahl (eds.) Variation within Optimality Theory: proceedings of the Stockholm workshop. University of Stockholm.

Bermúdez-Otero, Ricardo (2006) “Phonological change in Optimality Theoryin Brown, Keith (ed.) Encyclopedia of language and linguistics. Elsevier.

Bermúdez-Otero, Ricardo (2007) “Diachronic phonologyin de Lacy, Paul (ed.) The Cambridge Handbook of Phonology. Cambridge University Press.

Bermúdez-Otero, Ricardo (2011) “Cyclicityin van Oostendorp, Marc; Colin Ewen; Elizabeth Hume and Keren Rice (eds.) The Blackwell Companion to Phonology. Wiley-Blackwell.

Bermúdez-Otero, Ricardo (2013) “Amphichronic explanation and the life cycle of phonological processesin Honeybone, Patrick and Joseph C. Salmons (eds.) The Oxford Handbook of Historical Phonology. Oxford University Press.

Chomsky, Noam and Halle, Morris (1968) The sound pattern of English. New York: Harper and Row.

Jasanoff, Jay H. (2004) “Plus ça change… Lachmann’s Law in Penney, John H. W. (ed.) Indo-European perspectives — studies in honour of Anna Morpurgo Davies. Oxford University Press.

King, Robert D. (1968) Historical linguistics and generative grammar. Prentice Hall.

King, Robert D. (1973) “Rule insertionLanguage 49:551‒578.

Kiparsky, Paul (1965) Phonological change. PhD thesis, MIT.

Kiparsky, Paul (1968) “Linguistic universals and linguistic change.” in Bach, Emmon W., Robert T. Harms and Charles J. Fillmore (eds.) Universals in Linguistic Theory. London: Holt, Rinehart and Winston. pp. 170‒202.

Kiparsky, Paul (1973) “On comparative linguistics: the case of Grassman’s Law.” in Diachronic, Areal and Typological Linguistics, vol. 11 of Current Trends in Linguistics. Mouton. pp. 115‒134.

Kiparsky, Paul (1988) “Phonological change.” in Newmeyer, Frederick J. (ed.) Linguistics: the Cambridge survey. CUP.

Kiparsky, Paul (1995) “The phonological basis of sound change.” in Goldsmith, John A. (ed.) The handbook of phonological theory. Blackwell.

Kiparsky, Paul (2006) “The amphichronic program vs. Evolutionary PhonologyTheoretical Linguistics 32:217‒236.

Kiparsky, Paul (2008) “Universals constrain change; change results in typological Good, Jeff (ed.) Language universals and language change. OUP.

Ohala, John J. (1989) “ Sound change is drawn from a pool of synchronic Breivik, Lev Egil and Jahr, Ernst Håkon (eds.) Language change: contributions to the study of its causes. Mouton.

Ohala, John J. (1992) “What’s cognitive, what’s not, in sound Kellerman, Günter and Michael D. Morrissey (eds.) Diachrony within synchrony: language history and cognition. Peter Lang.

Ohala, John J. (1993) “The phonetics of sound change.” in Jones, Charles (ed.) Historical linguistics: problems and perspectives. Longman.

Roberts, Philip J. (2009) An Optimality-Theoretic Analysis of Lachmann’s Law. MPhil thesis, University of Oxford.

Roberts, Philip J. (2012) Towards a computer model of the historical phonology and morphology of Latin. DPhil thesis, University of Oxford.

Watkins, Calvert (1970) “A case of non-chronological rule insertionLinguistic Inquiry 1:525‒527.