September 21, 2020

Jeffrey Heath Interview – Historical Linguistics

I want to talknow about historical linguistics and the role they played in my career andsome thoughts about the right way and the wrong way to do historicallinguistics through the right way and the wrong way to look at it.

Historicallinguistics like any historical enterprise including political historyor social history or what-have-you has to focus ultimately on the issue whether the transmission of languages or social structures acrosstime is basically revolutionary or basically conservative.

Now the issue, thedifficulty in making that decision is that one way of.

.

.

one form of conservatism is keeping the basic underlying structure going while changing all thedetails so that the surface is changing but the the underlying substance is notthis is what makes historical linguistics ultimately I think aninteresting issue – do we change for the sake of change? Is change constantlyoccurring on its own? Is it is it driven by some external propulsion? or is itjust a lot of reshuffling of the cards so that you end up with the same hand atthe at the end of the day.

Alright I was an undergraduate at Harvard in 1967 to'71.

I had the very good fortune of being among some very good historicallinguists there, including Calvert Watkins who was probably the dominantIndo-Europeanist of his era including Frank Cross a specialist in Northwest Semitic languages including Thorkild Jacobsen a famous Assyriologist and also Einar Hagen a famous Norwegian linguist with a lot of interests inlanguage contact and heritage languages and language politics and things of thatsort in addition there were junior faculty members who were very helpful tome including Anthony Arlotto so it was an environment where there wasa lot of interest in historical linguistics there was alsosome interest in American Indian languages there.

There was in particularthere were people there working on Algonquian languages and languagesrelated to Algonquin including Karl Teeter and Ives Goddard so there was a circle of people and although I never really got into those particularlanguages or language families very deeply I absorbed a lot from thatenvironment and it had had a lot of influence on my thinking about languagein general and not just about processes of historical change.

In a nutshell my.

.

.

as my career has evolved my emphasis in historical linguistics has been oncycles of disruption and repair now by disruption I usually mean some kind ofphonological changes that are disruptive to the to the morphology and I will saythat most of the languages I've worked on are languages with fairly rich andsome cases extremely rich morphology so that keeping the system going is no easymatter when there are disruptive forces at work.

Now the disruptions themselvesare somewhat interesting.

Perhaps the least interesting or the most obvious isjust simple phonetic attrition of the sort that frequently happens at the endof words where the final vowel or the final consonant or even the finalsyllable gets reduced or deleted or things like syncope whichdoes pretty much the same thing immediately in words of three or moresyllables so just simple phonetic attrition which is it just it happensit's the the normal course of events over time for phonological systemsnow I became interested in one particular type of phonological shiftnot necessarily.

.

.

.

not necessarily destructive but more in terms ofshifting features.

I became interested in the idea that the vowel cycles thatpeople have talked about going back to the English vowel shift – the greatvowel shift and also the various cyclical phenomena that Bill Labov andothers have described for English dialects I became interested in the ideathat one driving factor in that is a a pull toward the sound the vowel [i]written with with the letter i usually in IPA but [i] as the sound andthat the movements of the type [æ] to [ɛ] to [e] to [i] were pull chains as there was something pulling people or certain types of people to that and I connectedthat with John Ohala a phonetician his work on frequency code and on theidea that the the vowel [i] because of its high second formant and its secondformant interacts with the third formant and so you get a kind of a.

.

.

what isperceived often it's a high-pitched vowel even if it isn't really high-pitched andthat that fed into the business about daintiness or lack of threatof things that are high-pitched and therefore you can be presumed to besmall and inoffensive as opposed to darker heavier vowels like [ɑ] which wewould associate in the other direction that are perceived tosome degree is having lower pitches and coming from larger creatures thanthe ones that have vowels like [i] so I became interested in that as a way ofaccounting for.

.

.

it's one of the drivers of these cyclical changes.

Now for thatto work you have to kind of get into delicate territory because thesocial linguistics of those things is this often described as being verygender specific with women moving in one direction and men either moving nowhereor moving occasionally in the other direction and there was a lot ofevidence for that – not that that's the that gender as such is the realphenomenon in question it's much more of a psychological thing or a.

.

what youmight call femininity and masculinity as symbolic patterns but anyway there did seem to be considerable evidence that this was going on sothat's one disruptive effect and I did a paper with Matt Gordon the sociolinguist Matt Gordon is now at University of Missouri on that incurrent anthropology.

Another type of in addition to just regular attrition andthis type of social linguistically motivated or social phoneticallymotivated shift all of which can disrupt by neutralizing by merging vowels and soforth in addition to that there's substratum effects in the course oflanguage shift.

So typically you've got a substratum they've got their languagewith their prosodic and phonological patterns and another languagecomes in, sweeps over and in a certain number of generations everybodyswitched to the new language the target language but they keep theprosodic and phonological patterns of the original substratum language, so thiswould be your your rapid spread of languages like Indo-European, which endedup being pronounced in radically different ways in different parts ofthe world depending on what the substrata were, Romance languages whichare very different in.

.

.

pronounced very differently in France and theIberian Peninsula and in Romania because of the languages the substrata languagesor even adstrata languages so the idea of language contact that hasphonological effects all of these can disrupt morphology a very good exampleof the the latter is old the oldest dialects of Moroccan Arabic which I amconvinced were lost their vowel length distinction under the influence of asubstratum which was none other than Latin None other than late Latin asstill spoken into the.

.

.

into perhaps the eighth century in the northern areasof Morocco.

So there are disruptive effects most of which don't haveanything to do with function.

.

.

they don't care about function they don't careabout what disruptions they caused in the morphology but once they happenthe morphology has to patch itself up has to keep going and so how does thathappen? well I have to go back to my undergraduate days at Harvard where myspecialty my own specialty was Uto-Aztecan languages.

Now the Boston area is not terribly well known for as a location where you're going toget lots of Hopi and Shoshone or Aztec people goingthrough.

So my work was entirely from books Fortunately Uto-Aztecan isblessed with a very long tradition of excellent descriptive grammars goingback to Edward Sapir even beyond that to some of the great Mexican Spanishlinguists of the post-colonial period and people like Benjamin Lee Whorf andSidney Lamb and a lot of other people have worked on individual Uto-Aztecan languages so there's there is a very good literature on many of them butwith the exception of the very early work by Edward Sapir on the historicallinguistics which is which was based on a very limited amount of data therehasn't been very much good historical work on Uto-Aztecan and so I had an opportunity to do that and I wrote a very long undergraduate honors thesisseveral hundred pages although that was with very large letters and generousmargins and generous line spacing so probably would have been much more thana 150 pages nowadays but it was my first major work in historicallinguistics.

Now it was about verbal morphology and verbal morphophonemics and it was I think competent but not theoretically very exciting but later onI became.

.

.

I went back to Uto-Aztecan and it was from the point of view ofhow repair functions when you have when you have attrition or otherprocesses that are making the the incumbent or inherited morphologyputting it on the edge of extinction so typically when you have suffixes thatare down to one consonant because the vowel has been attrited.

And so i was interested in the process of how those things are renewed And I showed that, or tried to show that in several cases what had happened was that they kept the word forum with the suffix at the end being attrited and what they did is they brought in anothermorpheme such as a verb that was hanging around out there and they brought that inand the connection being that the verb began with the same consonant as whatthe endangered suffix that was still there but just barely hanging onin the form of a single constant you brought it in so what you were doing wasyou were saving the morphology by bulking up on the suffixphonologically bulking up where it was in the process of disappearing.

.

.

hadn'tquite disappeared but was on the verge of disappearing.

Now I called thosehermit crab restructurings and I found some other examples of that.

I'm actuallyinclined to think that the Germanic dental preterites have a.

.

.

the development of that in Gothic I think is the same kind of process whereyou have you have a suffix and you have a verb out here and they get they getconnected so that the larger paradigm brings in a verb form inthe form of the suffix well anyway that was that was one form of repair wherethe key is that the suffix and the independently occurring stem begin withthe same segment in general with the consonant.

So there's a link notnecessarily a very close link in the meaning but a phonological connectingthread that facilitates that merger now this is an in opposition to modelswhere you when you see as a suffix that looks like a verb that you can identifywith the verb you assume that there was a an original construction, with one verb here in another verb here perhaps a subordinated verb or aninfinitive or something and that you have a gradual process of grammaticalization where this gets squashed into from syntax into morphology a kind ofthing that the grammaticalization theorists talk about.

So I was arguingfor a different process a more abrupt one where you repair an existingstructure rather than jettisoning it and allowing a new structure to come in fromthe syntax and become gradually grammaticalized.

Another example of this type thatI found in connection with Australian languages was what I called lost wax asa process of repair.

Now lost wax as a way of doing bronze sculpture.

What youdo is you create molds you have an inner mold and an outer mold and in betweenyou've got some wax that's in the shape of the thing you're trying to make andthen what you do is you pour molten bronze into the cracks and it drives outthe wax and then the bronze hardens and then you pull the molds out and you'vegot your structure.

The reason that that metaphor was interesting to me wasbecause in these Australian languages you have what we call a direct inversesystem what that means is that you have a subject and object marking in the verbsubject and object agreement you have but instead of having the subject herein its slot now all the different subject possibilities in another slotwhich has the object markers instead you have a linear order that's determined bypronominal hierarchy so first and second person always comes first andthen the third person human plural and the third person human singular and thenthe various non-human classes.

So you don't know just from the just from theposition of the pronominals which one is the subject in which one is theobject.

So what you do is you put a morphemecalled the inverse morpheme in between the subject and the.

.

.

in between the firstelement and the second one if the first element is the object and the second oneis the is the subject.

So if it's “me” and then “him” here – if that means “I saw him”or I did something to him nothing happens if it means he did something tome it's gonna be “I” inverse “he” and thenthe rest of the of the verb.

So this is a critically important relational morpheme.

This is a morpheme that you can't do without because if it disappears youhave massive confusion as to what is subject and what is objectwell in some of these languages the inverse morpheme was in fact attritedand it was in some cases it was a nasal and then it would be zeroed beforeanother nasal and so that was really a problem that was looming and what thelanguage did was it found another morpheme that had a syllable shape andoccurred as a kind of epenthetic context and a very limited range offorms on the edges of this inverse system and it got a got a footin the door in one of these combinations and then it spread like wildfire throughall of the of the inverse combinations.

So you end up with a very productive.

.

.

so you saved the system while replacing the key the absolutely essentialmorpheme being replaced throughout by this other morpheme and this must havehappened before the old morphine was totally dead must have happened whilethe old morphine was still visibly present and at least a subset of thosecombinations.

So it was a repair It was not something were you the systemcollapsed and then you started all over again and you rebuilt it up from scratch.

It was a repair it was a it showed how important it was for speakers of thisimage to keep that system going even under the threat of attrition.

So that'sanother another way to do it not involving phonological.

.

the same consonants or anything like that but the same basic principle.

Now in the Moroccan casethat I'm talking about, if you know anything about Arabic you know thatvowel length is very important grammatically and lexically so forexample you make a participle out of the verb by lengthening the the the [a] vowelyou various plurals are expressed by with long vowels it's very important inthe in the morphology as well as in the basic lexicon.

Now when Moroccan Arabic first came into northern Morocco and you had peoplespeaking late latin and you had only a few Arabised Berber soldiers holding thefort for the Arabs you had a situation where the substratum pronunciation coulddramatically affect the pronunciation of arabic and that's whathappened and at the time late Latin like early Romance had long since lost vowellength distinctions from from early classical Latin and so the vowel lengthdisappeared in the oldest forms of Moroccan Arabic.

Now that sets off awhole set of changes particularly in verbs.

I won't go into the details butit led to a dramatic restructuring where what you end up with is a kind ofcrazy identity of one half of the perfected paradigm with the imperfectiveparadigm leaving the other part of the protective paradigm by itself.

We'retalking about the form of the verb itself before pronominal endings andit's because of the flattening of the distinction between short [o] and long[oː] and short [i] and long [iː] and the result is to createartificially or secondarily or by accident a homophonybetween the imperfective and part of the perfective paradigm and then thispattern spreads to other cases where there was no historical reason for thatchange to happen – just that became the pattern that the imperfective forall categories and the first and second person perfective that's now amorphological unit and then the third person perfective is a different one andthat spreads throughout the verbal system so this is not so much a repairbecause it doesn't really.

.

.

it doesn't really solve any functional problems but it's an example of a reorganization that is triggered – accidentally – by aphonological system anyway I think you can gather from the comments I'vemade that I don't have much use for extreme forms of grammaticalization theory I think that in many cases of what's really happening is notthat you have a syntactic construction that has a kind of built-in destiny tocontract and move into the grammar and push some some older incumbentconstruction out of the grammar without any reference to what's around it I'mmuch more interested in historical processes that are conservative innature, rather than unnecessarily disruptive in nature.

I'm talking aboutgrammatical changes not phonological and where the the system the incumbentsystem brings things in selectively to keep itself going so it doesn't justget bombarded with new constructions barging their way into the grammar, rather the native system the speakers of the language have a lotinvested in that native system and they they work hard to keep it going.

So my general philosophical approach is thathistorical linguistics with the exception of the language contactphenomena but if languages are left on their own without socially disruptivephenomena happening involving other languages they're gonna.

.

.

they'regoing to be fairly similar from one point in time to another.

Even whena lot of the the actual grammatical morphemes have been reshuffledand so I think I understand what Edward Sapir was getting at when he thought he said long ago that languages the pattern of a particular language family could bedistinctive even when the morphemes have been replaced by other morphemes.

He wasthinking of cases like Algonquin and the the California languages that arerelated to it.

He was thinking of things like Athabascan and he was looking atthe he was thinking about how you get these patterns which keep going even asthe superficial details vary.

Now recently.

.

.

I used to think that alot of this was about functionalism I used to think that repair was somethingbecause when you had these disruptive forces it meant that there was goingto be ambiguities and so forth and that was really what the repair wasall about.

I have lately begun reconsidering that, although I stillthink that where there are.

.

.

there is a threat of significantambiguity that that's that is enough to to induce speakers to repair the grammarbut I've been reading all this work about simplification and incomplexification of languages and I've been reading the work of people likelike McWhorter and so if worth and they've partially convinced methat elaborate morphological complexity is kind of decorative it isn't reallynecessary for conveying messages that speakers are.

.

.

also David Gill points outthat you don't really need much grammar to get your meanings across as long asthe speaker is coming from the same place that you are and that has made mereconsider the idea that repair in historical linguistics is primarilyabout saving the ship from sinking It might simply be at least in some casesor to some extent that it's just really a matter of habit.

That is that speakersnew speakers come into the world and they learn the language of their oftheir parents and their elders and they acquire habits of speaking and that asas these phonological disruptive effects happen they disrupt the way you havelearned to speak they don't necessarily cause massivemiscommunication or ambiguity that gets you into trouble or causes people todrive off cliffs and so forth but that it.

.

.

it is.

.

.

just disrupts the way youare used to speaking and the people around you have been speaking so that itmight be that repair is something that's going to happen even with forms ofmorphology that aren't really critically essential but that are just the waywe've.

.

.

we are used to speaking and that would be a different kind of way oflooking at at conservatism that it isn't just about fighting tomaintain the integrity of language as a communicative system – but it's it's a wayof fighting to maintain the the styles of speaking that we areused to.

Leave a Reply

Your email address will not be published. Required fields are marked *