Introduction :  

            The first hybrid languages to attract sustained linguistic attention were the pidgin/creole languages of the Caribbean region.  The academic interest generated by the Caribbean linguistic scenario led, initially, to an interest in identifying analogous language situations elsewhere in the world and ultimately, in the absence of a more comprehensive overview of hybrid languages, to an implicit conclusion that pidgin/creole languages were the most important hybrid language phenomenon.  Simplification, convergence and other more general participating processes in language hybridization were, as a result, investigated primarily in the context of these pidgin/creole situations.  This explains the current tendence to regard all simplified, converged, or otherwise hybrid languages, at least initially, as plausible instances of pidginization or creolization.

            When pidgin/creole theory penetrated Africa-proper and looked beyond its European-related pidgins and Creoles, it encountered a socio-linguistic situation far more complex than it was designed to comprehend.  Whereas in the Caribbean terms like `simplification', `convergence', `pidginization', `creolization' and `hybridization' could be used more or less interchangeably to refer to the local situation, in Africa the pidgin/creole situation was simply one among many hybrid language types showing, outwardly, simplification and convergence.  Nor was it clear how to label, let alone demarcate, these various linguistic phenomena on the basis of current pidgin/creole-based theory.  It is clear, however, that the Caribbean was limited and therefore arbitrary as a starting point or for any objective investigation of language hybridization in all its complexity.

            India , like Africa , presents a range of language situations complex enough to test current theoretical assumptions about hybridization processes and develop a more comprehensive theory.  But the theory itself must initially be coherent and integrated enough to allow a precise identification of each hybrid case taking into account the full range of possibilities.  This means first being able to anticipate the potential importance of a given hybrid phenomenon in India .  Existing theory must then be made clear and explicit before it can productively be put to test.  Only then can empirical research provide the final test, which will confirm or refute our speculation.

PIDGINS.

            The label `pidgin' has casually been applied to many language situations in India where simplification and convergence are involved.  But how likely is true pidginization in a language environment like India ?

            The classic pidgin situation involves two socially segregated communities, where the segregation is based on a perceived ethnic/class distinctness.  One of the communities, termed the `super stratum', is socially superior and/or alien, and linguistically more or less homogeneous.  The other community, termed the `sub stratum', is usually menial in status and/or a conquered people.  The sub stratum community is linguistically heterogeneous to the extent that while its languages are genetically and typologically very close.  Between the super stratum language and the sub stratum languages there exists a genetic and typological gulf which is too wide for either group to have reliable linguistic intuitions about the structure of the other's language(s).

            The new presence of the super stratum community is responsible for bringing into permanent social contact, the diverse, ethnic groups which make up the sub stratum community, thus creating a new social environment.

            The need for a new, pidgin language arises from this brand new social environment where, within the sub stratum, one new stable ethnic group has been created where earlier, there existed several smaller groups, each with its own language.  The need for a link language between the super and sub stratum communities is a very minor, marginal consideration, and indeed, the super stratum community generally has to learn pidgin consciously from the sub stratum, and usually speaks it imperfectly.

            In the formation of the classic pidgin, the sub stratum community resorts to its common `specific linguistic sub-stratum' as a general model for the pidgin structure.  The specific linguistic substratum is an abstract construct which takes into account the typological similarities between the languages of the sub stratum, which the pidgin makers unconsciously discover working on the basis of each smaller group's linguistic intuitions and concept of normalcy.  Above this level of commonality, where surface conventions are more likely to be in conflict, structural simplification discards most of the surface complexity.  Where structural compromise seems impossible, the pidgin may incorporate and adapt super stratum structures, or optimize its structure on the model of language universals.  The lexicon is the most conspicuous area of structure where compromise is impossible, being also the most arbitrary area.  Thus, pidgins draw and adapt the bulk of their lexicon from the super stratum language.

            The resulting pidgin is not to be confused with the initial inter language forays between the super stratum community and a small part of the sub stratum group, where unstable, adhoc language mixture and simplification exist in plenty.  Individual ad hoc attempts at communication in such situations are instances of "Incipient bilingualism" and not pidginization: since the contact is at a direct individual level, the sub stratum speaker here can ultimately learn the super stratum language, and this is generally his goal.  However, these `incipient bilinguals' do provide the channel whereby super stratum structure and lexicon reach their larger group, cut off from the super stratum language by segregation, but seeking to form an internal link language of its own.

            In a variant pidgin scenario, the sub stratum community is made up of groups, or individuals, whose original linguistic backgrounds are so diverse that there is no specific linguistic substratum to which they can refer in order to shape their pidgin.  Their original languages are as different from each other as they are from the super stratum.  These situations produce unstable pidgins, where performance and, indeed, the grammatical framework as a whole differs from speaker to speaker.  Unlike the classic pidgin, which is stabilized early by the pidgin-makers themselves, the variant pidgin type, termed the `universal substratum pidgin', is stabilized only in creolization by the universal grammar competence of the children in the next generation, who acquire it as a nation language.

            When contact between the super and sub stratum group does not result in a permanent and total disruption of the social order for the sub stratum groups, a stable pidgin may persist, immune to creolization, as long as the contact lasts.  The sub stratum groups, in this case, retain their original languages as native, and maintain their original ethnic distinctness despite the addition of a new social link between the groups.

            When super and sub stratum contact does result in permanent and total social disruption (particularly relocation) for the sub stratum groups, the creation of the new group generally leads to an erosion of internal ethnic distinctions, and the next generation linguistically reflects this new homogeneity with the pidgin as its native language.  A creole is simple a pidgin turned native language, and expanded in structure and function by the children who speak it natively.  When a creole is further affected in structure by a constant magnetic pull from its super stratum language, it becomes a post-creole.  Creoles and post-creoles, unlike pidgins, resemble `natural languages' in their degree of structural expansion.

INDIA .

            A large number of non-standard and non-mainstream languages in Idia are regularly referred to as `pidgins' or `pidginized languages'.  These languages all, in some measure, seem `simplified' and/or `mixed' in structure when compared with the official standard languages.  But this on its own is not sufficient grounds for determining prior pidgiization.  For a language to be a pidgin it has to have originated and evolved within the specific `life-cycle' we have defined as the process of pidginization.  To include other simplified or mixed languages as pidgins is to make the term `pidgin' meaningless.  Before we consider the more superficial pidgin-like features in these languages, then, the language situations themselves must first be evaluated against our abstract model of the process of pidginization.  Unless we use an abstract model as our guide we run the risk of including as pidgin languages which only look like pidgins outwardly or, at best, identifying a small number of the pidgins in existence, but only those with a close outward resemblance.

            It is difficult to assess `structural simplicity' in objective terms on the basis of purely synchronic information.  Many features of non-standard (and indeed, standard) languages which seem `simple' in comparison with analogous areas of closely related languages may not be simplification at all, but rather independent developments which have nothing to do with prior hybridization processes.  It is not unusual for non-standard languages in rural and urban India to have less elaborated surface structure than the related prestige languages, and if the term `language universal' has any meaning outside the context of hybridization, one has to expect this particularly in a restricted code.  Comparison with related standards, in such situations, is both in valid and misleading.  Non-standard and non-mainstream languages, far from being pidgin or any other sort of recent hybrid development, often have longer histories as stable languages that do the recognized social standards.  It is not inconsistent for as a simple-looking language to be fully a `natural' language in terms of its recent history.  It is more than likely that most of India 's non-standard/non mainstream languages are healthy "natural languages".

            Non-standard and, more particularly, non-mainstream languages often do not remain `healthy' when faced with drastically changing social situation, where the speech community gradually adopts a new language for more and more of its daily functions.  This situation usually arises when an outside language extends its area of hegemony, encompassing new linguistic territory and creating new language loyalties among its people (as in India 's tribal areas), or, more dramatically, when an entire speech community shifts to a new location within alien linguistic territory.  This can produce various results.  The entire community may abandon its ethnic language within a generation and be absorbed linguistically into the new local setting (as with the Kashmiri pundits in the Hindi/Urdu heartland).  A community may alternatively maintain its ethnic language fully along side the local language, particularly if its ethnic language is a prestige language and the relocation is viewed as not necessarily permanent (as with Begalis in north-eastern India).  However, less elite communities often have less unequivocal responses to this type of linguistic trauma.

            When a language is abandoned by its speech community, gradually or abruptly, this phenomenon is called `language death'.  Language death is a process of reduction, potentially affecting three aspects of a language: the size of the speaker community, the number of functions discharged by the language, and the actual structure of the language.  Of these three, it is the process of structural reduction which elicits comparison with pidginization.

            When language structure is reduced in a language death situation this outwardly reflects a shifting linguistic `centre of gravity' within a bilingual community where, at the one extreme, speakers' ethnic and native languages are one and the same, while at the other, speakers' native language is the new language of the area.  This makes the process as a whole a complex phenomenon wherein, at the one extreme, structural loss is in terms of features perceived as `expendable', and entirely consistent with existing tendencies in the language, while at the other, the language as a whole becomes a restricted code used for special purposes by speakers without native-speaker intuition: a situation not unlike pidginization.  There are, however, important reasons why this cannot be considered pidginization.

            Firstly, pidginization is not a gradual loss of structure but rather the prompt emergence of a new, simple language for a group with no other language in common.  In a language death situation, it is the healthy presence of a second community language which has weakened the earlier ethnic language, hence this restricted code no longer serves its original unifying function.  Instead, as with incipient bilingualism, it provides an exogenous link: to an older generation.

            However, there is no a prior reason why a given language death situation should culminate in a total loss.  Indeed, after an initial period of structural loss, the process can be arrested at any point in the spectrum and the restricted code re-nativized and even re-expanded in structure in response to a change in the social climate.  This appears to have happened to relocated languages spoken by less elite communities in various parts of India .  The renativization process obscures the earlier language less leaving, on the face of it, an overtly `mixed', `simplified' language ethnically unifying a bilingual community.  A language lexically related to the earlier ethnic language, but structurally not incompatible with the (perhaps homogeneous) local situation, though generally based on the earlier ethnic language.  While such a hybrid greatly resembles a creole, it is too shallow a mixture, and too discerningly simplified, to be a true pidgin/creole.

            Indian bazaar vernaculars are often another area of suspected pidginization.  Some of these languages are clearly not pidgins at all: our definitions automatically exclude from this category all individual ad hoc codes which do not reflect the behavior of a single coherent community.  These cases overwhelmingly tend to be temporary interlangauge phenomena associated with incipient bilingualism, used for communication outside the social group.  Such situations do not produce new, viable languages.  When individuals gain access to a new language through anew contact with social equals who speak the language natively, for example in migration to a major city, they generally learn the new language properly after some initial discomfort.  There is no social barrier to prevent this.

            Other cases, however, are less clear-cut.  Imagine a spectrum if hybrid types, ranging from the simplified compromise variety spoken at a regional market, where many villages with closely related dialects participate, to the more fundamentally `mixed' variety spoken by a heterogeneous group of labourers, with a lexicon largely adapted from the language of the managerial group.  The first scenario is koine, a stable, simplified `everyman's language', almost equidistant from the various source dialects, unifying a larger social group.  The second, however, recalls the Caribbean plantation, recreated in an Indian factory town or railway town.  The only essential difference from the Caribbean scenario is that the underlying specific substratum on which the hybrid is based usually takes account of the grammatical structure of the super stratum language as well, since most of the languages in contact are genetically and typologically close.

            Such a hybrid, then, differs from a koine and, indeed, a pidgin only in terms of degree.  This discovery, in turn, raises the exciting question as to whether koines and pidgins are in fact related along a smooth conceptual specutrum, which one could calibrate in terms of the degree of genetic/typological distance between the super and the sub stratum language inputs.  At the pidgin end of the spectrum there may even exist, in heterogeneous areas of India , like the tribal north-east, hybrid languages wit the requisite typological gulf between the super and sub stratum languages to allow full pidginization.  There is anecdotal evidence of such developments in Arunachal Pradesh which supports this speculation.

            The koine-pidgin spectrum offers the most feasible model for investigating the development of Urdu in various parts of India .  If we consider its north-Indian beginning we find an alien power group entering and settling in a heterogeneous dialect situation where no prestige/standard language already exists.  Their entry and assumption of power changes the social climate of the area bringing a new feeling of cohesion.  In the army barracks and the urban areas a new language emerges among the indigenous people: zaban-e-urdu, the camp language.  The power elite retain as their language Persian.

            While this history meets all the social conditions of pidginization, the linguistic structure of the hybrid Urdu suggests a far less traumatic origin.  The lexicon is certainly no less Indic than Persian, and the grammatical structure is almost exclusively Indic, and hardly simplified at all.  Persian, though genetically related to the North-Indian dialects, is typologically dissimilar enough to play the distant super stratum language of the classic pidgin situation.  Why, then, did classic pidginization not take place?

            The answer appears to lie in the nature of the earlier North-Indian dialect situation.  While admittedly no recognized standard language existed in the area before the Moghul presence, the smaller dialects themselves were not so lexically distant as to be mutually unintelligible to any significant degree.  So, even though no official standard variety had emerged, there must have been a tradition of communication between the groups using koine varieties whose functional role was more limited.  At the very least, there must have been a tradition of ad hoc link language formation based on the clearly perceptible level of grammatical and lexical commonality between the various dialects.

            Indeed, this exact situation currently obtains in the Bhojpur-speaking area of India , where, in the absence of perceived cohesion, and with Hindi language loyalty among the upper classes, no recognized standard variety has emerged.  Even so, there is adequate mutual intelligibility between the regional dialects, and there appear to exist unofficial koine varieties for wider communication.  Given such a tradition, then, where sis the need for a new alian lexicon?  And given the closeness of the original dialects, and the comparative shallowness of the specific substratum, where is the need for radical simplification to achieve compromise?  So the new group language, Urdu, has an essentially North-Indian grammatical structure, and its core lexicon is North-Indian.  Persian lexicon either labels phenomena associated with the new social-religious presence or is used variably, as a stylistic option.

            In other areas of India , notably Hyderabad and Madras , Urdu has a hybrid character more reminiscent of the pidgin/creole situation.  Among the less elite converted Muslims, Urdu is a mixed language, displaying South-Indian substratum features incorporated into the grammatical structure, with a North-Indian Urdu lexicon.  Here, too, in response to a new and dominant presence, a new social grouping has formed within the indigenous population which, however, remains socially segregated from the power elite. 

            The new group language, however, has been adopted less because of any total communication impasse than with an aim to differentiate the new group from the rest of the wider society.

            It is difficult to place such an hybrid on any koine-pidgin spectrum without more precise information about the initial contact situation, particularly information on the extent of linguistic heterogeneity among the would-be converts, and the extent of their social interaction with the power group.  Indeed, natural languages (and earlier koines) do not have to undergo hybridization to be adopted by new populations.  The task here, then, is to assess how substantiative the South-Indian element is in the structure of these southern Urdu varieties and, piecing together the original contact situation, to discern whether the simplification and mixture in Southern Urdu is radical enough to be compatible with our notions of pidginization.  If, so, its subsequent nativization would make it a creole.

            It is important to be cautious about concluding pidginization here, since striking parallels also exist between southern Urdu and another neo-Indian language derivative of a colonial situation: Indian English.  Indian English is distinctively Indian and spans a continuum ranging, at the basilect, from radically simplified varieties with a completely Indian syntax (for example, an SOV word order rather than SVO, postpositions rather prepositions), to acrolectal levels which differ from International English primarily by an overuse of infrequently found English sentence types (such as passives: `this is required by me', `this is to be given for fixing', and statives: `this is necessary'(=I need this), `this is available' (= you can get this)).  These acrolectal structures are not inconsistent with English grammar to the extent that they cannot be classed as actual grammatical errors, but they are Indian in conception.  They are adaptations of normal Indic sentence types.  But despite this evidence of simplification, mixture, plus a sociolinguistic continuum fraught with mesolectal traps and claques, Indian English is not a pidgin/creole, or any other sort of hybrid.

            Unlike the Moghuls, the British never sought to penetrate or reform Indian society, creating new social groups.  They were simply an addition, a top layer, as it were, over a stable social situation.  Individual Indians acquired English purely in order to communicate with the British, and learnt it was well as each contact situation permitted.  This created a diglossic situation, with English sharing place with the earlier ethnic languages.  Indian English did not originate at a single coherent basilectal point from a single stable hybrid.  Its very origin was a serious of individual events, compatible with various points along the formal continuum, as largely unrelated instances of incipient bilingualism.  Only later did it serve any function of uniting a recognized indigenous group, such as the bureaucrats, and even then, only in a restricted way, since in most situations outside the office the unifying language even for this group was indigenous.

            Present-day Indian English, however pervasive and distinctive it is, however much of a social phenomenon it may have become, is not a pidgin/creole.  Yet it shares with the Caribbean post-creole situation a system of socially based constraints on the extent of proximity a given speaker might ultimately expect vis-à-vis the standard variety.  But the vast inter language area between the pre-basilectal Indic vernaculars and the foreign standard has no necessary relationship to the pidgin/creole situation: indeed, a largely shared lexicon between two distinct languages (related or not) is the only linguistic prerequisite for a post-creole type continuum.  There is no social reason why such a continuum could not develop between two discrete Indic languages, such as Hindi and Bhojpuri which, partly for socio-political reasons, largely share a lexicon in much the same way as do basilectal Creole English and Standard English, though Bhojpuri, unlike Hindi, has a Magadhan-type structure.  Neither is a hybrid development of the other.

            What overwhelms the imagination in the case of Indian English is simply the massive scale on which incipient bilingualism has occurred.  A situation not unlike French linguistic hegemony in Romania and Imperial Russia, where a foreign language, besides its exogenous communicative value, attains the status of a class marker, effectively dispossessing the bulk of the population.  Indian English, too, without even being an interesting hybrid, is a daily source of defeat for India 's bourgeoisie, and unreachable for the lower strata of the population.  And an intellectual sinecure for the elites.  But the magnitude of this `incipient bilingualism' should not surprise one in a country like India , where a tiny percentage of the population amounts to a staggering number of people.  Indian English probably touches no more than five percent f the population in any significant way.

CONCLUSION.

            Perhaps the best evidence of the ultimate `naturalness' of nativized hybrid languages, and of the existence of `universal grammar' as a linguistic lodestar, is the way hybrid and non-hybrid languages can grow to be functionally and structurally so alike as to present problems in determining which socio-linguistic phenomenon we are confronted with.  But before we can hope for an explanatory theory which could reveal an underlying `unit of all the forces' we have to clarify in precise detail what the individual processes are within the broader sociolinguistic panorama.  This we could never do on the basis of the Caribbean setting alone: it has been revealed as too unique and monolithic to offer the range of situations we need to develop our theory.  India and Africa are better for this.  But empirical initiative must take into account that the theory is still in a stage of plasticity, and may well be rejected by the data.  Hence, every Sociolinguist must also be a theorist, or risk misunderstanding his own data.

            This is extremely important because the social environments we envisage for the future seem less favourable to the development of extreme pidgin languages.  Slave plantation societies no longer exist, and the more overt forms of colonialism are on the decline.  Ethnic groups newly in contact are normally less overtly segregated than they would have been in the past.  But present-day languages are increasingly in contact with each other, larger social groups are forming with new language loyalties, and many smaller languages are being abandoned.  So our theory is not only currently confronted with more complex situations in India and Africa , but with situations which promise to become even ore complex and less pidgin in the future.  We have to be sensitive to these environmental demands on our theory, or run the risk of becoming irrelevant in the future.

REFERENCES :  

Bickerton, Derek, 1975.

            `Creolization, linguistic universals, natural seman-tax and the brain'.  Read at the International Conference on Pidgins and Creoles, University of Hawali .  

Duran.  James J. 1979.

            `Non-standard forms of Swahili in West-Central Kenya ', In Hancock (ed.) 1979: 129-151.  

Ferguson , Charles A. 1959.

            `Diglossia'. Word 15.  

Gilman, Carles, 1979.

            `Commeroonian Pidgin English, a neo-African language'.  In Hancock (ed.) 1979: 269-280.  

Givón, Talmy. 1979.

            `Prolegomena to any sane creoogy'.

            In Hancock (ed.) 1979: 3-35.  

Hancock, Ian F. (ed.). 1979.

            Readings in Creole Studies.

            Ghent : Story-Scientia.  

Mohan, Peggy. 1976.

            `To coin a koiné theory'.  Read at the conference on New Directions in Creole Studies, University of Guyana .  

. . . . . . . 1981.  Review of Readings in Creole Studies, ed. Ian Hancock. To appear in Language.  

Samarin, William. 1971.

            `Salient and substative pidginization'.

            Pidginization and creolization of languages, ed. Dell, Hyemes. Cambridge : Cambridge University Press.