INTRODUCTION
The period of European colonial expansion, starting in the 1500s, saw the transplantation of a number of European languages to other continents. The languages most involved in this process were Portuguese, Spanish, French, Dutch, and English. In the fullness of time, this transplantation led to the development of a number of new national and local varieties of these languages, such that they were all clearly different from varieties in the metropolitan homeland.
Now if we ask why new varieties of these languages developed in the new locations, then we can cite a number of different factors, such as linguistic change, adaptation, and language contact. But it seems obvious that dialect contact and dialect mixture must also have been very important factors in determining the nature of colonial varieties of European languages, such as South American Spanish, Brazilian Portuguese, Afrikaans, Canadian French, and the colonial Englishes. Surely if you take English speakers from all over the British Isles and settle them in a single location on, say, the east coast of Australia, dialect mixture will be the inevitable result.
In fact, it emerges that not everybody has always been happy to accept this rather obvious fact (see below). Dialect contact itself would appear to be accepted as a given, in view of the fact that in nearly all cases European settlers arrived from many different locations in the metropolitan homeland. But what does not seem to have been a given for some historical linguists is that dialect contact necessarily leads to dialect mixture and thus to new-dialect formation. Indeed, a number of them have rejected the importance of dialect contact in the development of European colonial dialects altogether and plumped for monogenetic origins for particular colonial varieties.
To take some of them in chronological order: Rivard 1914 suggested that Quebec French is entirely due to the dialects of Normandy. Wagner 1920 was only one of a number of hispanicists who produced the andalucista theory that Latin American Spanish is in origin basically a form of transported Andalusian Spanish from the Iberian Peninsula. For English, Wall 1938 claimed that New Zealand English was simply transplanted Cockney. Hammarström 1980 made the same Cockney-origin claim for Australian English. Laurie Bauer has also argued more recently that “it is clear that New Zealand English derives from a variety of English spoken in the south-east of England” (Bauer 1994:391). He goes on to argue further that New Zealand English arrived as a ready-formed entity transplanted in its entirety from Australia: “The hypothesis that New Zealand English is derived from Australian English is the one which explains most about the linguistic situation in New Zealand” (1994:428). These two points are of course entirely consistent if one accepts that Australian English, too, arrived from the southeast of England. And Lass 1990 has argued that South African English is essentially southeast of England in origin.
Perhaps one should not be too surprised if the role of dialect contact in leading to dialect mixture has been rejected by such writers. Dialect mixture is in many ways a more subtle phenomenon, and one that is more difficult to detect, than language contact; and, especially, it is also a good deal more mysterious in terms of its origins.
It is really very clear why various forms of language contact should have the consequences that they do. But why exactly should dialect contact result in dialect mixture? It is not surprising if some linguists reject the one as necessarily leading to the other – because why should it? Why should speakers adopt features from dialects other than their own – something which obviously has to happen if mixture is to occur? If we define dialect contact as contact between language varieties that are mutually intelligible, then why would speakers modify their behavior at all in the presence of speakers of other dialects who are able to understand them perfectly well even if they do not modify?
In fact, there are some explanations for this. The hypothesis which I advanced in Trudgill 1986, which seems to have received some acceptance (e.g., Tuten 2003), is that the fundamental mechanism leading to dialect mixture is accommodation in face-to-face interaction. Accommodation in face-to-face interaction is a concept developed by Giles 1973 and further refined in other publications such as Giles, Coupland & Coupland 1991. Tuten, in his brilliant book, says that “given that most contributing varieties in a prekoine linguistic pool are mutually intelligible … many of the alterations in speech that take place are not strictly speaking necessary to fulfil communicative needs.” He continues: “Trudgill's emphasis on accommodation reveals rather novel assumptions about why dialect contact leads to change” (Tuten 2003:29). And it emerges that he agrees with my “novel assumptions.”
But actually, of course, an acceptance of the role of dialect mixture, and thus of accommodation, simply leads to yet another question. Why is it, in fact, that speakers do accommodate to each other in face-to-face interaction? Why does that happen?
Tuten has an answer:
Given that most contributing varieties in a prekoine linguistic pool are mutually intelligible … many of the alterations in speech that take place are not strictly speaking necessary to fulfil communicative needs … Rather, speakers accommodate to the speech of their interlocutors in order to promote a sense of common identity. (Tuten 2003:29; my emphasis)
Here I part company with Tuten. Although there clearly are sociolinguistic situations where identity plays a role, I see no role for identity factors in colonial new-dialect formation, and I have particular trouble with Tuten's use of the phrase “in order to.” But I have to acknowledge that this sort of view about the motivation for accommodation and thus for the development of new colonial dialects is rather widely held. For example, the Australian lexicographer Bruce Moore (1999) has said, of the development of colonial Englishes: “With language one of the most significant markers of national identity, it's not surprising that post-colonial societies like Australia, the United States, Canada, New Zealand, should want to distinguish their language from that of the mother tongue.”
Schneider, in his article on new varieties of English (2003:238), also says of the development of these varieties by colonial “settlers in a foreign land” that “the stages and strands of this process are ultimately caused by … reconstructions of group identities.”
And Macaulay (2002:239) says: “I fully expect new dialects to develop in places where a sense of local identity becomes strong enough to create deep-seated loyalty.”
Perhaps most importantly, Hickey (2003:215) has written in a critique of Trudgill, Gordon, Lewis & Maclagan 2000 that New Zealand English is to “be seen as a product of unconscious choices made across a broad front in a new society to create a distinct linguistic identity.” He then goes on to argue that the selection of certain variants available in a dialect mixture “can be interpreted as motivated by speakers' gradual awareness of an embryonic variety of the immigrants' language, something which correlates with the distinctive profile of the new society which is speaking this variety.”
In my view, we would do well to be a bit more skeptical than this about explanations for the formation of new colonial varieties couched in terms of identity. I share the kind of view expressed by Mufwene when he writes (2001:212) of the development of creoles that they were not “created” by their speakers but that they emerged “by accident.” I share, too, the skepticism expressed by William Labov on the importance of identity factors in leading to linguistic change. Labov's famous Martha's Vineyard study (1963) is often cited as a typical and very telling example of the important function of identity in producing language change. Strikingly, however, Labov himself does not agree. He writes:
The Martha's Vineyard study is frequently cited as a demonstration of the importance of the concept of local identity in the motivation of linguistic change. However, we do not often find correlations between degrees of local identification and the progress of sound change. (Labov 2001:191)
So what sort of evidence might there be that I could bring forward to support my skepticism about the thesis that issues of new national identity have played any role at all in the initial development of colonial varieties of European languages? I suggest that one of the clearest pieces of evidence in favor of my view is the fact that dialect mixture and new-dialect formation must have been of considerable importance for the development of new colonial varieties of European languages well before the period of European overseas colonial expansion that began in the 16th century. The point is that dialect mixture resulting from accommodation in dialect contact situations played a role in determining the nature of new colonial varieties of European languages at many periods of language history which predate the colonial expansion period, and at times when, and in places where, new, national colonial identity factors are most unlikely to have played an influential role.
A number of examples follow. Note that I do not produce any positive evidence concerning the absence of any role for identity factors. I do not believe that that would be possible. Rather, my evidence is negative, which is why I make the argument at some length and on the basis of as many as five cases where, I suggest, no evidence in favor of the role of identity can be found.
COLONIAL MIDDLE ENGLISH IN IRELAND
The first place that English was exported to overseas was Ireland. It did not arrive in the Americas until the 1600s, while it arrived in Ireland centuries before that. My focus here is on late medieval English in Ireland as studied by Michael Samuels. Samuels (1972:108) shows that the colonial English of Ireland in the 14th and 15th centuries was clearly the result not only of dialect contact but of actual dialect mixture. First, he says that the dialect forms found in the available texts show that the English settlers “must have been predominantly from the West Midlands and South-West England.” Importantly, however, the language of the texts “tallies with the dialect of no single restricted area of England; it consists mainly of an amalgam of selected features from the different dialects of a number of areas: Herefordshire, Gloucestershire, Somerset, Devon, Shropshire, and to a lesser extent Cheshire, Lancashire, and possibly Wales.” Most of the forms are found in large parts of the West Midlands and Southwest, but they do not all have the same geographical provenance. For example, streynth ‘strength’ and
‘through’ are from the West Midlands, while hyre ‘hear’ and ham ‘them’ are Southwestern in origin.
He also focuses on a particular linguistic feature which is of special interest because it clearly constitutes a feature of the type I have referred to as an “interdialect” feature. In Trudgill 1986 I introduced the term interdialect to refer to forms of a number of different types which are not actually present in any of the dialects contributing to a dialect mixture but which arise out of interaction between them. Samuels focuses on a type which I labeled intermediate forms. He shows that in England, the Middle English form for ‘each’ was uch in the West Midlands, and ech in the Southwest. However, in the Anglo-Irish texts the norm is euch. It is true that there is a small area in England, intermediate between the West Midlands and the Southwest, just to the south of the towns of Worcester and Hereford, where this form also occurs. But the fact that it is the norm in Anglo-Irish texts is of course not because Ireland was settled from southern Herefordshire. Rather, euch was a form which developed in Ireland, or survived from England, or both, because it was intermediate between the two major competing forms in the mixture that developed in Ireland, uch and ech.
It is clear, then, not only that the relevant areas of Ireland were settled by English speakers from a number of different dialect areas, but also that the outcome was a new, mixed dialect, consisting of a collection of forms from different dialect areas, plus interdialect forms arising out of interaction between speakers of different dialects. And I suggest that we would surely be struggling if we wished to argue that, at this period of history, this new, distinctive form of Irish English was in any way developed as a consequence of the medieval anglophone settlers in Ireland wishing to signal some kind of national identity separate from that of English speakers in England.
GERMAN COLONIAL DIALECT MIXTURE
English is, of course, not the only language to have been subject to pre-16th-century colonial expansion. An early proponent of the importance of dialect mixture in the formation of new colonial dialects was Frings 1957, who, on the basis of an analysis of the German Linguistic Atlas materials, and of historical records, described the large-scale eastward colonial expansion of German into relatively less populated, formerly Slavic-speaking areas as leading to dialect contact and dialect mixture.
Up until about 900 the border between Germanic and Slavic-speaking areas ran from the Baltic coast along the river Elbe as far south as Magdedurg and then along its southwestern tributary, the Saale, down as far as Hof. Thereafter, the “Ostkolonisation” saw the frontier between German and Slavic (and Baltic)-speaking peoples move eastwards, from 900 to 1350, into what is now eastern Germany and then beyond, into what is now Poland and parts of modern Lithuania.
The initial movement of this “east colonization” was across the river Saale to the middle Elbe. There were three major colonization routes. The first was from the Netherlands and northern Germany via Magdeburg and Leipzig (Slavic and Modern Polish Lipsk). The second was from central Germany via Erfurt and Leipzig. And the third was from southern Germany via Bamberg and Chemnitz (Slavic Kamjenica, Modern Polish Kamienica). Settlers arriving via the first route brought with them northern Low German dialects; the second stream brought Central German dialects, and the third Bavarian and other South German dialects. These three routes converged in what is now Saxony, the main focus of the convergence being Meissen (Slavic Misni, Modern Polish Miśnia), which had been founded as a German city in 929. Meissen then formed the main jumping-off point for German-speaking colonization farther east to Dresden (Slavic Drezdany, Modern Polish Drezno), which was a German city by 1206, and on into Silesia and beyond.
It was in the area of Meissen that, according to Frings, a new colonial dialect developed during the 12th and 13th centuries. Because of the mixture of settlers from different areas, a new “kolonial Ausgleichsprache” was formed. Frings described the result as being a mixture of Dutch, Low German, Central German and Upper German dialect forms. (He also argued that this was later to provide something of a basis for written Standard German, but that is another issue.)
This 14th-century Meissen “Ausgleichsprache” shows a mixture of regional forms from different areas to the west. These include northern he ‘he’ as opposed to central and southern er; and northern dit, dat ‘this, that’ as opposed to central and southern dies, das; the central German diminutive -chen rather than northern -kin or southern -lein; central Fund ‘pound’ as opposed to northern Pund, southern Pfund; the central and southern pronominal distinction between accusative and dative forms, for example mich ‘me (acc.)’ and mir ‘(to) me (dat.)’, which is absent from northern German; southern Ochsen rather than central and northern Ossen; and southern diphthongization in Haus as opposed to central and northern Hus (see especially Frings 1956:3:20).
We can ask if the mixing and subsequent focusing which took place in the newly Germanized lands, and which led to the development of a whole new mixed dialect, took place as a consequence of any factors connected with some new colonial identity. Certainly we have no evidence that this was so, and it seems that the issue of German versus Slavic identity would have been of much greater importance.
COLONIAL DIALECT MIXTURE IN THE FORMATION OF IBERIAN SPANISH
Similarly, Spanish too was subject to a form of colonial expansion, as described by Tuten 2003 himself, long before the settlement of the Americas, which also had important linguistic consequences. Along the northern edge of Iberia, from the Atlantic via the southern edge of the Pyrenees to the Mediterranean, there was, and still is, a dialect continuum with considerable linguistic variation and small dialect areas. Traveling from west to east, one encounters the Ibero-Romance varieties Galician, Asturian, Castilian, Aragonese, and Catalan. However, as one travels south toward the southern coastal areas of Andalusia and Murcia, the degree of dialectal differentiation diminishes and dialect areas become bigger. This is the result of dialect mixture processes which occurred during the reconquest of the peninsula by Ibero-Romance-speaking Christians, who took over from the Arabic and Mozarabic and Berber-speaking Muslim Moors.
As Penny 1991, 2000 pointed out, the southward Iberian expansion and the dialect mixture that went with it account for a number of historical developments in the history of Spanish, and therefore for a number of the characteristics of the modern language. According to Penny's thesis, there were three major phases in the development of Spanish during which dialect mixture led to koineization, and then rekoineization of already koineized varieties. As Tuten 2003 describes it, the first phase of southward colonial expansion lasted from the 9th century to the 11th and focused on the city of Burgos, in north central Spain, where Ibero-Romance speakers from all over northern Spain – but especially northern Castile, Asturias, Navarre, and Leon – came together as part of the southward expansion. The second phase of koineization took place between the 11th and 13th centuries and focused on the central city of Toledo. Here again, dialect mixture occurred on a large scale. Finally, the third phase, which took place between the 13th and 14th centuries, focused on the southwestern city of Seville. Once again, considerable dialect mixture occurred as populations speaking different dialects moved in from the north to replace the Moors.
To cite just one relatively straightforward linguistic example of the consequence of these mixtures, the disappearance from Castilian of contracted forms of prepositions plus definite articles, found in many Romance languages, is one of the most dramatic changes in the history of Spanish, and yet, Tuten says, it has been little discussed and never explained. His explanation is that in the mixture associated with southward expansion during the Burgos period, there would have been a large number of variants of, for example, en + los, such as Galician nos, Leonese enos, Aragonese ennos, and Riojan and northern Castilian enos. The Burgos post-koineization Castilian forms which emerged from the mixture were the simpler and more analytical, such as en los. It is well known that dialect contact can lead to regularization and more analytic structures (Trudgill 1986), and Tuten suggests that during koineization speakers would have selected or developed forms which were more easily analyzed and generated and “whose component parts also appeared regularly in other contexts” (2003:119).
In other words, the Iberian Spanish varieties which were taken westward across the Atlantic, and which underwent dialect mixture processes there, were themselves already the result of several earlier stages of mixing and koineization. Three different new koines were formed in Spain at three different periods in three different places, as a result of colonial expansion, dialect contact, accommodation, dialect mixture, and new-dialect formation. But there seems no reason to suppose that issues of new colonial identity were involved in their formation. On the contrary, the colonial expansions each took an already existing non-Arab, non-Muslim, Christian, and later Spanish identity southward with them.
VALENCIAN AS COLONIAL CATALAN
A very similar kind of process appears to have happened in the case of Catalan. As a result of the Christian reconquest from the Moors, Catalan spread southward from its original homeland in the northeast of the Iberian Peninsula into the province of Valencia (Baldinger 1958, Ferrando 1989). In the military struggle which led to the Christian victory, the Aragonese were predominant but the Catalans made significant contributions of money, men, and the ships required for provisioning the king's army, making landings, and blockading the coast (Bishko 1975). The annexation was led by the Aragonese-Catalan monarch James I from his capital in Saragossa. The military campaign began in 1232 and ended in 1245, when James gained control of the east coast area right down to the borders of Murcia, which, it had been agreed, was going to Castile.
Once again we can suppose that dialect contact occurred as the newly available lands were settled by incomers from the north, although the area of the northern Iberian dialect continuum from which the incomers came was much shorter geographically than in the case of Castilian, and we can therefore expect fewer dialects to have been involved. The major Romance varieties that penetrated into Valencia are usually agreed to have been, from west to east, Aragonese; Western Catalan, from the area around Lleida; and Eastern Catalan, from the area around Barcelona.
In his chapter “La constitució del vocalisme catalá” Alarcos (1983:57–78) says that the outcome of this mixture was such that the new Valencian variety of the language came to resemble West Catalan. However, he argues that this is a kind of coincidence. Valencian resembled West Catalan but was not West Catalan, just as Canadian French is not Norman, South American Spanish not Andalusian, and Australian English not Cockney (Trudgill 2004), all of them being results of dialect contact and dialect mixture. For example, Alarcos (1983:75) claims that some aspects of the Valencian vowel system are “fàcilment explicable com un fenomen d'anivellament entre diversos estrats lingüístics simultanis.”
A rather complex series of different changes in different places to the Western Vulgar Latin vowel system led to a situation where the Old Eastern Catalan three-way distinction /e/,
, and
–
‘within’ ∼
‘grass’ ∼
‘land’ – corresponded to an Aragonese two-way distinction between /e/ and /ie/:
BUT
and /tiera/ and a differently distributed Western Catalan two-way distinction between /e/ and
:
,
BUT
.
Then, “en el necessari procés d'igualació entre uns parlants i altres, devien abandonar-se aquelles articulacions que només fes servir un dels grups” (1983:75). The forms which were abandoned were thus /ie/ and
, giving rise to the Valencian two-way distinction between /e/ and
:
∼ /erba/ but
, which, as Alarcos says, just happens to be the same as that of Western Catalan in this case.
A similar new-dialect formation phenomenon occurred in the case of word-final unstressed /a/, /e/, which had been merged on
in Old Eastern Catalan, but maintained as distinct both in Aragonese and Western Catalan. Old Eastern Catalan
:
-
corresponded to both Aragonese /a/ and Aragonese /e/ –
-
– as well as to Western Catalan /a/ and /e/,
-
. The majority form won, giving rise to non-merged Valencian
-
, which again just happened to be identical with Western Catalan. Alarcos says: “El valencià, doncs, coincideix amb el dialects occidental pel simple fet d'haver sorgit per un procés d'anivellament dialectal,” in which the variants common to the most varieties were “triumphant.” Although we can agree that a separate Valencian identity is important to many people in the area today, there is not much reason to suppose that this is anything except a fairly recent phenomenon, and no reason at all to suppose that it predated 13th- and 14th-century new-dialect formation.
OLD ENGLISH AS A COLONIAL MIXED DIALECT
The Danish linguist Hans Frede Nielsen has argued that the earliest example of English colonial dialect mixture leading to new-dialect formation involves the actual development of English itself. We have evidence of various nonlinguistic sorts that southern and eastern England and southeastern Scotland were initially settled by Germanic speakers coming from all along the North Sea littoral from Jutland to the mouth of the Rhine – Jutes, Angles, Saxons, and Frisians – and it is reasonable to assume that they spoke different dialects. We also know that Germanic speakers also arrived from further inland, for example from place-name evidence. Swaffham in Norfolk, for instance, was originally Swaefasham, which meant ‘home of the Swabians’, who were a non-coastal people. But did contact between these different Germanic dialects lead to dialect mixture? Can we say that Old English was in origin a mixed dialect, just as we are claiming for, say, modern South African English?
Nielsen (1998:78–79) answers this question in the affirmative, and supplies the linguistic evidence. He argues that Old English was the result of a mixture of West Germanic dialects from continental Europe. He points out, for example, that it is because of dialect mixture that Old English had initially a greater degree of variability than the other Germanic languages where no colonial dialect mixture had been involved. He cites the following examples.
(i) Old English (OE) had a remarkable number of different, alternating forms corresponding to Modern English ‘first’. This variability, moreover, would appear to be linked to origins in different dialects from the European mainland: ærest (cf. Old High German eristo); forma (cf. Old Frisian forma); formesta (cf. Gothic frumists); and fyrst (cf. Old Norse fyrstr).
(ii) Similarly, OE had two different paradigms for the present tense of the verb to be, one apparently resembling Old Norse and Gothic, and the other Old Saxon and Old High German:
(iii) Old English also exhibited variation, in all regions, in the form of the interrogative pronoun meaning ‘which of two’. The form
relates to Gothic hvaþar and W. Norse
while the alternative form hweder corresponds to O. Saxon
and OHG hwedar.
The suggestion, then, is that even if we did not know from other nonlinguistic evidence that southern and eastern Britain were initially settled from many different locations on the Continent, there would have been at least some linguistic evidence that would have pointed us in the direction of that supposition. We can thus suppose that Old English was a new colonial variety of West Germanic that resulted from a mixture of continental dialects. But can we suppose that the speakers who produced this new variety were motivated in any way by a sense of new colonial, quasi-national identity? It seems unlikely.
CONCLUSION
If the development of these earlier colonial varieties cannot be seen as having been motivated by the development of new national colonial identities, then the uniformitarian principle (Labov 1994:22), which states that “knowledge of processes that operated in the past can be inferred by observing ongoing processes in the present,” leads us to suppose that the same was true of the later and better-known post-16th-century colonial expansion period also. I thus claim that new mixed colonial varieties can and do come into being without identity factors having any involvement at all. We do not need this as an explanatory factor at any moment in human history.
Of course, since the heyday of European colonialism, new identities most certainly have developed in most of the colonies. French Canadians are no longer French; Australians are certainly not British; Afrikaners are very definitely not Dutch; and these new identities do have a strong linguistic component. But my suggestion is that if a common identity is promoted through language, then this happens as a consequence of accommodation; it is not its driving force. Identity is not a powerful enough driving force to account for the emergence of new, mixed dialects by accommodation. It is parasitic upon accommodation, and is chronologically subsequent to it. Identity factors cannot lead to the development of new linguistic features, and it would be ludicrous to suggest that Australian English speakers deliberately developed, say, wide diphthongs in order to symbolize some kind of local or national Australian identity. This is, of course, not necessarily the same thing as saying that once new linguistic features have developed they cannot become emblematic, although it is as well to be skeptical about the extent to which this sort of phenomenon does actually occur.
Let us consult Labov again. His view is that, before one jumps to conclusions based on notions of identity, patterns of interaction should always be consulted for possible explanations. Labov's main preoccupation in his writings on this topic has been with the diffusion of linguistic forms; but new-dialect formation, which depends just as much as diffusion on how individual speakers behave linguistically in face-to-face interaction, can be regarded in precisely the same way. Labov argues that “as always, it is good practice to consider first the simpler and more mechanical view that social structure affects linguistic output through changes in frequency of interaction” (2001:506). He bases his argument on Bloomfield's assertion that:
every speaker is constantly adapting his speech-habits to those of his interlocutors … The inhabitants of a settlement … talk much more to each other than to persons who live elsewhere. When any innovation in the way of speaking spreads over a district, the limit of this spread is sure to be along some lines of weakness in the network of oral communication. (Bloomfield 1933:476)
Labov argues that it follows from this that “a large part of the problem of explaining the diffusion of linguistic change is reduced to a simple calculation” (2001:19). It is purely a matter of who interacts most often with whom – a matter of density of communication. Labov then develops the principle of density:
The principle of density implicitly asserts that we do not have to search for a motivating force behind the diffusion of linguistic change. The effect is mechanical and inevitable; the implicit assumption is that social evaluation and attitudes play a minor role. (Labov 2001:20)
But why, we can ask, is it “mechanical and inevitable?” What exactly is inevitable about it?
I do have an answer to this question that I wish to propose. The answer is that it is inevitable because accommodation is not only a subconscious but also a deeply automatic process. It is, as I have argued in Trudgill 2004, the result of the fact that all human beings operate linguistically according to a powerful and very general maxim. Keller (1994:100) renders this maxim as “Talk like the others talk.”
Keller's maxim, in turn, is the linguistic aspect of a much more general and seemingly universal (and therefore presumably innate) human tendency to “behavioral coordination,” “behavioral congruence,” “mutual adaptation,” or “interactional synchrony,” as it is variously called in the literature. This is an apparently biologically given drive to behave as one's peers do.
There is a copious literature on this topic (see, e.g., Cappella 1981, 1996, 1997; Bernieri & Rosenthal 1991; Burgoon, Stern & Dillman 1995), which suggests that linguistic accommodation is not driven by social factors such as identity at all but is an automatic consequence of interaction. Pelech (2002:9), for example, says that “the innate biological basis of interactional synchrony has been established.” He then goes on to say that “the ability to establish interactional synchrony represents an innate human capacity and one of the earliest forms of human communication,” and that this capacity served, and serves, “the basic survival needs of bonding … safety, and comfort.” Cappella 1981 explores further the evolutionary and biological bases for the existence of adaptation processes in the human species. And in Cappella (1997:65) he says that “mutual adaptation is pervasive” and that it is “arguably the essential characteristic of every interpersonal interaction.” Linguistic diffusion and new-dialect formation are “mechanical and inevitable” because linguistic accommodation is automatic, because, as Cappella (1997:69) says, it is an aspect of “the relatively automatic behaviors manifested during social interaction.”
My suggestion is therefore that it is this innate tendency to behavioral coordination, not identity, that is the very powerful drive that makes dialect mixture an almost inevitable consequence of dialect contact, to an extent that factors connected with identity would not and could not. And it is also this drive that led to the development of new, mixed colonial varieties of European languages. The actual linguistic characteristics of any new mixed dialect result from the relatively deterministic principles outlined in Trudgill 2004; and it is the new mixed dialect to which the founder principle – that the speech of the founding population of a colony determines what its dialect will be like (Mufwene 2001:28) – then applies.
In fact, there is an increasing body of evidence that suggests that large-scale and prolonged dialect contact always leads to dialect mixture, and therefore in a sense requires no explanation, and certainly not one in terms of identity. It is true that Cappella says that the behavior involved in accommodation is only “relatively” automatic, the implication being that it can be overridden in the case of individuals. But behavioral congruence is the default; and prolonged large-scale dialect contact will always lead inevitably to dialect mixture and to new-dialect formation.
