Historical Linguistics Research Paper

1. The Scope Of Historical Linguistics

Historical linguistics involves the study of languages as they change through time. It includes tracing developments from an earlier stage of a language to a later one: for instance, English ‘silly’ meant ‘blessed’ at the time of Chaucer; Latin sentences often had the word order Subject Object Verb, as in Helena Marcum amat ‘Helena loves Marcus,’ whereas its daughter language French typically has Subject Verb Object, as in Helene aime Marc; and many Old English noun and verb endings with final nasal consonants, like -an, -um, -on, dropped gradually about 1,000 years ago. Historical linguists are interested in when, how, and ultimately why changes of this sort happened.

However, this diachronic perspective also operates in the opposite direction, encouraging historical linguists to look back from present-day or recorded languages to earlier, unattested stages in an attempt to reconstruct unrecorded features and characteristics. This may involve filling in gaps in the historical record, for instance in hypothesizing the form a word might have taken in Old English or in Latin, although that word does not actually appear in any surviving manuscript. At greater time depths, features of unattested, hypothetical ancestral languages, like Proto-Indo-European, the probable source of most languages of modern Europe and northern India, may also be reconstructed. Work of this kind also explicitly or implicitly involves consideration of language classification: reconstruction necessarily includes the hypothesis and testing of linguistic family groups, which are taken to be demonstrably descended from a single common source. At present, it is uncontroversial to say that English, French, and Latin are all related, as members of the Indo-European family; but it would be more controversial to contend, as some historical linguists do, that these languages are in turn related to Semitic languages like Arabic, Akkadian, and Hebrew; or that Chinese is related to Chechen and other Caucasian languages.

In this research paper, the development of historical linguistics will first be considered briefly, and its place within modern linguistics assessed. Then the two subdisciplines of language change, and linguistic classification and reconstruction, will be examined in turn, with the emphasis on outlining currently lively and productive areas of research. In the final section, some trends common to change and reconstruction will be identified, with their implications for future work in historical linguistics.

2. Historical Linguistics Within Linguistics

In the heyday of traditional work in the late nineteenth century, historical-comparative linguistics was linguistic theory. Its focus and methodology were in line with the strongly historicizing tendency of the day, and with the development of parallel sciences such as geology and evolutionary biology, so that although the foundations for language comparison and study had been laid considerably earlier, the discipline advanced rapidly and attracted considerable scholarly attention in this period (see Davies 1998). The cornerstones of the discipline were set with the classification and reconstruction of Indo-European in particular, and the Comparative Method on which such reconstructions are based (see Fox 1995) is still the foremost tool of the reconstructionist’s trade, although many of the specific results obtained have been modified, as knowledge of language change has deepened and developed, and more data have become available. The Comparative Method requires sets of cognate forms in descendant languages: these will be similar in form and meaning, and will show regular, recurrent correspondences of particular sounds across the group of languages compared. We can then hypothesize that these have been derived historically from a single most plausible common source, by the operation of generally regular sound changes. Thus, given English father, Irish athair, Latin pater and Greek pater, alongside a large number of similar comparands which equally lack an initial consonant in Irish, and have English /f/ corresponding to Latin p , it seems reasonable to hypothesise Proto-Indo-European *p ter. It is also possible to set out the contexts in which particular sound changes have operated in the daughter languages, and therefore to begin to identify the factors in the surrounding sounds which may have led to these developments.

Although many of these methodological and theoretical assumptions still hold good in historical linguistics today, the twentieth century has seen a radical shift in perspective in linguistics as a whole. The development of structuralism led to a move away from the study of individual languages and families, and towards identifying language universals, those aspects of human language which recur regardless of genetic linguistic affiliation. Consequently, the historical linguistic concentration on variation as a key to linguistic relatedness gave way to attempts to identify theoretical limits on variation. Finally, whereas most earlier work had concentrated on vocabulary and phonology, the ‘Chomskyan revolution’ of the 1960s centered attention on syntax.

However, these rapid and far-reaching changes have not led to the demise of historical linguistics; although it is true that historical work became less of a core part of linguistics for some years in the mid-twentieth century, it is now again one of the liveliest areas of the discipline. This is for three main reasons. First, advances in linguistic theory, while typically based on synchronic study and data, often allow the sources of change to be identified and explained in a way which was not previously possible: much nineteenth-century work, although allowing individual elements to be plausibly reconstructed, was piecemeal and did not trace related changes to a common cause. Explanation comes center stage in modern linguistic work. Second historical data provide an additional means of testing and refining these linguistic theories, which may benefit from a demonstration that they can analyze and incorporate language change. Finally, related disciplines like genetics and archaeology may not always find developments in linguistic theory per se particularly useful or enlightening; on the other hand, interactions of data and hypotheses from these areas with those of historical linguistics are promising. All these areas will be considered, with respect to both change and reconstruction, in the following sections.

3. Areas Of Debate In Language Change

3.1 Phonology

Much early work on sound change was based on the assumption of relative ease of articulation: that is, it is presumed to be more straightforward for a speaker to utter two similar adjacent sounds, or indeed to make them identical, than to produce a sequence of sounds which require major vocal tract movements between the two. This explanation for the extremely common sound change of assimilation, as seen in Latin somnum → Italian sonno, was put forward in the last century. However, advances in phonetic technology has allowed historical linguists to pinpoint more accurately the types of sound sequences which are likely to cause such problems, and the advent of experimental phonology has allowed these hypotheses to be tested, not only from the point of view of the speaker’s production problems, but also taking the hearer’s perception difficulties into account. One of the pioneers in this area, John Ohala (1993), has suggested that certain sound changes like dissimilation, which increases the difference between close or adjacent sounds in the stream of speech, may be provoked by the hearer’s expectations. Thus, a hearer receiving the ProtoQuechumaran word */t’ant’a/ ‘bread’ might erroneously have assumed that the speaker has produced two ejective [t’]s close together by assimilating a second, nonejective stop to the first truly ejective one. Undoing this perceived but not real mistake by the speaker produces the modern Quechua form /t’anta. Advances in phonological theory may also be relevant here, for instance in helping us understand what it means for sounds a and b to be closer in their constituent features than sounds b and c; or in showing that the apparently disparate environments triggering a particular sound change in fact form a unified class in terms of some deep principle of phonology. In this respect it is encouraging that, while certain phonological theories current in the 1980s, such as Government Phonology and some forms of Declarative Phonology, explicitly ruled historical processes out of their remit, Optimality Theory, a currently popular model which has been developing since the early 1990s, already includes a considerable body of work addressing problems of both change and variation.

3.2 Morphology And Syntax

Morphology has often been left out of historical linguistics as a separate domain of enquiry, being seen as simply an interface: phonological changes, like the loss of final unstressed syllables, may have morphological consequences if case or tense markers are lost, and this in turn may have a knock-on effect on the syntax, changing or fixing word order. However, morphology has more recently been the focus of attempts within Natural Morphology (Wurzel 1989) to identify those features which are favored universally, for instance in terms of ease of production, perception, or acquisition, and the factors underlying that naturalness, as well as assessing why in many cases these universally natural morphological features are subordinated to language-specific requirements which may reduce naturalness. Morphological change is also one component of grammaticalization, the extremely common process whereby independent words become grammatical markers, as in German Drittel ‘third’ and Viertel ‘quarter,’ formed from the numerals drei ‘three’ and ier ‘four’ plus a reduced form of Teil ‘part’; or the famous French negative ne … pas, where pas is derived from Latin passum ‘a step,’ and became an integral part of the negative construction in earlier French, perhaps via gradual bleaching of reinforcing constructions meaning ‘I won’t go a step.’ Grammaticalization has become a major focus of research in linguistic theory (see Hopper and Traugott 1993), introducing many students and researchers to issues in linguistic change.

In syntax, historical work has become enormously more sophisticated over the last 30 years, reflecting the newly ascendant status of syntax in synchronic work and in linguistic theory. Work at first may have been hampered by lack of knowledge of the range of syntactic changes, and uncertainty over quite fundamental questions, such as whether syntactic change can be seen as essentially regular in the same way as many sound changes. These issues are now being addressed by major cross-linguistic surveys (see Harris and Campbell 1995). On a more theoretical level, Lightfoot (1991, 1999) uses historical data and current models of syntactic theory in a mutually informing way, attempting to show that children learn syntax in the way they do because of particular factors in their innate, universal grammar, but that certain superficial changes in syntactic structures may also cause them to learn differently, setting up new structures which will form the basis for subsequent development in their language.

3.3 Semantics

Developments in linguistic theory may also increase understanding of semantic change, which has long been seen as a domain where description is the highest achievable goal. Earlier approaches have attempted to classify changes, but seeing developments as, for instance, extensions, or restrictions of meaning does not necessarily explain why these occur, and linguists have tended to fall back on the old adage that ‘every word has its own history.’ Developments in pragmatics and in cognitive linguistics may improve prospects here. For instance, Traugott (1982) suggests that many words develop from an external to an internal, subjective sense: thus, English feel begins as descriptive of some characteristic external to the speaker, in It felt soft, but comes to signal internal states, in I feel miserable. Sweetser (1990, p. 2) suggests that linguists must crucially adopt ‘a cognitively based theory which takes not the objective ‘‘real world,’’ but human perception and understanding of the world to be the basis for the structure of human language’: this approach allows recurring semantic changes, such as those involving mind-as-body metaphors of the grasp, cast light on, and see type, to be unified.

3.4 Variation And Change

One of the most productive recent developments in historical linguistics has been the recognition of a connection between variation and change, and the development of sociohistorical research. First, this substantially increases the database of changes available for study, since changes in progress can now be considered alongside the more traditional category of change from earlier, written data to the present day. However, for spoken, colloquial data to be accepted, it is necessary to delimit and define what constitutes a change in progress, requiring new quantitative techniques for identifying and measuring ongoing change. Innovations, which may provide the seeds of future developments, must now be distinguished from changes, and innovators from early adopters, the speakers who begin to use a new variant and spread it to their own social group (see Milroy 1992, Labov 1994).

This sociohistorical work has also directed attention away in part from internal, systemic factors and towards external factors in the search for explanations for change. Internal factors may belong to the language system itself, or to some other level of the grammar, or involve the innate language faculty which many linguists see as unique to humans; such factors have typically been favored by theoretical linguists in their attempts to explain change. However, sociolinguistic work places great emphasis on the social characteristics of speaker and interlocutor, including age, sex, and social class, and on characteristics of the interaction itself, such as level of formality and topic of exchange. This in turn encourages focus, not only on interaction between speakers, but between varieties and languages. The consequences of language contact have therefore become vitally important in recent historical linguistics, with increasing interest in contact phenomena like borrowing and convergence; the development of pidgins and creoles; and sociolinguistically determined language breakdown and loss in language death (Bickerton 1981, Thomason and Kaufman 1988, Dorian 1989). These are no longer being seen as peripheral and uncommon, but as a central part of language change and germane to a greater understanding of language itself.

4. Areas Of Debate In Reconstruction And Classification

4.1 Typology And Reconstruction

In the closely related subfields of classification and reconstruction, matters have advanced considerably since the various nineteenth-century attempts to write a fable in Proto-Indo-European. There is still much interest in what is reconstructed: the reliability and realism of reconstructed systems remain open questions, as does the issue of whether syntax can be reconstructed using the same methods as phonology, or whether early, unattested areas such as word order are beyond recovery (Lass 1997). Innovations in phonological reconstruction continue to occur: for instance, the ‘new look’ of Proto-Indo-European proposes a set of ejectives alongside the more traditional voiceless and voiced stops, on the basis that the resulting system would be more common cross-linguistically, even though the alleged ejectives have disappeared, largely without trace, in the daughter languages. This opens a wider debate over the roles of typology and uniformitarianism in reconstruction: that is, to what extent should reconstructions be driven by generalizations that hold for modern-day languages? Is it more important to reconstruct a proto-form or proto-system which accords with the comparative evidence from its daughters, but which might appear skewed when compared with other attested systems, or to produce a balanced, typologically normal system although considerable changes may be required to derive the daughter forms?

4.2 Classification And Methodology

However, classification has recently threatened to overtake reconstruction as the main focus of interest in this area, and here the questions are primarily methodological. The Comparative Method, with its concentration on regular, repeated correspondences and its incorporation, at least for many practitioners, of reconstruction as a test of the family groups proposed, has been challenged by Greenberg’s much simpler method of multilateral comparison, based on straightforward surface inspection of word-lists from many languages. Family groups are then proposed on the basis of etymologies drawn from these lists: forms within an etymology are similar to an extent in form and meaning, but the similarities need not be recurrent, and reconstruction is not necessarily undertaken. These families tend to be considerably larger (and hence potentially older) than those conventionally reconstructed using the Comparative Method. For instance, Greenberg (1987) uses multilateral comparison to propose only three language families for the Americas, including Amerind, a phylum or superfamily containing more than 1,000 languages. Even more controversial connections are drawn by other practitioners of multilateral comparison, for instance between the Sino-Tibetan and Caucasian languages; and there have even been attempts to produce universal etymologies, including forms from all or most language superfamilies, which might suggest forms of Proto-World, the alleged first language of mankind (see Ruhlen 1994).

4.3 Connections With Archaeology And Genetics

Multilateral comparison remains extremely controversial. Its apparent ease of use may be deceptive, not compensating for problems of reliability of data, and potential laxness of the criteria governing what counts as a phonetic or semantic match. Many historical linguists are reluctant to accept the method or its results, and therefore doubt whether language groupings above the traditional family level, and beyond a time threshold of perhaps 5,000–8,000 years, can be conclusively established.

Long-range comparativists are, consequently, increasingly turning to data from outside linguistics itself in an attempt to validate their work, and exploration of the potential for historical linguistics and related disciplines like archaeology and genetics to produce mutually informing and enforcing data is becoming central. For instance, Greenberg et al. (1986) cite genetic and archaeological data which seem to favour a three-way migration of peoples into the Americas, mirroring Greenberg’s tripartite classification of New World linguistic groups. Cavalli-Sforza et al. (1994), in work which caught the popular scientific imagination, analyzed genetic data from 142 indigenous populations worldwide, and argued that the resulting population tree closely matched a linguistic family tree including controversial linguistic phyla.

However, these correlations are themselves problematic. The lack of a one-to-one correspondence between language and archaeological culture means interpretation is all important, and increases the need for data from the fields under comparison to stand independently and to be subject to rigorous validation within their own areas before cross-disciplinary conclusions are drawn. In genetics, in particular, advances in the data-types which can be considered may produce a range of possible population histories rather than a single definitive answer: thus, Cavalli-Sforza et al.’s (1994) picture of the Americas, based on classical analysis of proteins, does not agree with the genetic groupings derived from more recent analysis of mitochondrial DNA and Y chromosomes; these will therefore produce population trees which need not match a particular linguistic tree.

Connections of historical linguistics with genetics and archaeology are also relevant beyond the domain of long-range comparison and influences move in both directions. For instance, Dixon (1997) applies the biological model of punctuated equilibrium to language, arguing that Comparative Method and the associated family tree model may be relevant only to periods of concentrated change and expansion in the histories of language families. At intervening periods of relative equilibrium, language contact and gradual diffusion of features would determine the course of language change. Geneticists and archaeologists are also increasingly seeking a historical linguistic methodology which is truly independent and repeatable, and which might transcend arguments among its practitioners, for comparison with their own data: the development of quantitative techniques of language comparison therefore remains a pressing goal within historical linguistics.

4.4 The Evolution Of Language

If our attempts to reconstruct might be blocked beyond a certain time-depth on methodological grounds, some progress might be made by starting from the other end: that is, by considering the origin of language in our species. This is a topic which logically belongs with historical linguistics, at least from a nonspecialist point-of-view, but which has traditionally been excluded from the field, being famously banned by the Societe de Linguistique de Paris in the 1860s. However, some rehabilitation has taken place during the 1990s, with increasing interchanges of ideas between historical linguists, computational linguists, primatologists, neurobiologists, and psychologists; Hurford, Studdert-Kennedy and Knight (1998), for instance, includes contributions on the possible emergence of phonology and syntax, with some use of computer simulations to show how learning of these new structures might have taken place, as well as work invoking more social and anthropological considerations to explain the rise of human language. Again, this is by definition interdisciplinary research, which raises questions fundamental to historical linguistics (such as how the, probably simple, initial constructions of early language developed into the modern range of variants), and to linguistics in general (such as the limits which might be set on what aspects of language are innately specified and which might be learned, this in turn depending in part on whether the underlying systems responsible for language evolved by natural selection or are in some way maladaptive).

5. Common Trends In Reconstruction And Change

In the sections above, change and reconstruction have been presented separately for convenience. However, these areas are not truly independent, and some common themes can be discerned which may be important for the future of historical linguistics in general.

First, it must be emphasized that change and reconstruction influence one another. For instance, if an early text or unknown language is deciphered, the database usable for drawing conclusions about changes is increased, but the same data may also confirm or refute the results of work on reconstruction.

Second, both subdisciplines need to become reconciled to the development and use of quantitative, computational approaches, and of tools. Work of this kind is probably more advanced in change, where techniques derived from sociolinguistics are already well established in some domains. The need is more pressing, and the issues involved more controversial, in classification.

Finally, it is encouraging to note that historical linguistics is becoming a major focus of research within linguistics. This means that historical linguists in turn must be willing to work with synchronically based theories, and to consider what cognate disciplines like genetics and archaeology might require, as well as what they can provide.

Bibliography:

1. Bickerton D 1981 Roots of Language. Karoma, Ann Arbor, MI
2. Campbell L 1998 Historical Linguistics: An Introduction. Edinburgh University Press, Edinburgh, UK
3. Cavalli-Sforza L L, Menozzi P, Piazza A 1994 The History and Geography of Human Genes. Princeton University Press, Princeton, NJ
4. Davies A M 1998 Nineteenth century linguistics. In: Lepschy G (ed.) History of Linguistics. Longman, London, Vol. 4
5. Dixon R M W 1997 The Rise and Fall of Languages. Cambridge University Press, Cambridge, UK
6. Dorian N (ed.) 1989 Investigating Obsolescence: Studies in Language Contraction and Death. Cambridge University Press, Cambridge, UK
7. Fox A 1995 Linguistic Reconstruction: An Introduction to Theory and Method. Oxford University Press, Oxford, UK
8. Greenberg J 1987 Language in the Americas. Stanford University Press, Stanford, CA
9. Greenberg J, Turner C G, Zegura S L 1986 The settlement of the Americas: A comparison of the linguistic, dental and genetic evidence. Current Anthropology 27: 477–97
10. Harris A C, Campbell L 1995 Historical Syntax in Cross-Linguistic Perspective. Cambridge University Press, Cambridge, UK
11. Hopper P J, Traugott E C 1993 Grammaticalization. Cambridge University Press, Cambridge, UK
12. Hurford J R, Studdert-Kennedy M, Knight C (eds.) 1998 Approaches to the Evolution of Language. Cambridge University Press, Cambridge, UK
13. Labov W 1994 Principles of Linguistic Change. I: Internal Factors. Blackwell, Oxford, UK
14. Lass R 1997 Historical Linguistics and Language Change. Cambridge University Press, Cambridge, UK
15. Lightfoot D 1991 How to Set Parameters: Arguments from Language Change. MIT Press, Cambridge, MA
16. Lightfoot D 1999 The Development of Language: Acquisition, Change and Evolution. Blackwell, Oxford, UK
17. McMahon A M S 1994 Understanding Language Change. Cambridge University Press, Cambridge, UK
18. Milroy J 1992 Linguistic Variation and Change. Blackwell, Oxford, UK
19. Ohala J 1993 The phonetics of sound change. In: Jones C (ed.) Historical Linguistics: Problems and Perspectives. Longman, London, pp. 237–78
20. Ruhlen M 1994 On the Origin of Languages: Studies in Linguistic Taxonomy. Stanford University Press, Stanford, CA
21. Sweetser E 1990 From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge University Press, Cambridge, UK
22. Thomason S G, Kaufman T 1988 Language Contact, Creolization and Genetic Linguistics. University of California Press, Berkeley, CA
23. Trask R L 1996 Historical Linguistics. Arnold, London
24. Traugott E C 1982 From prepositional to textual and expressive meanings: Some semantic-pragmatic aspects of grammaticalization. In: Lehmann W P, Malkiel Y (eds.) Perspectives on Historical Linguistics. Benjamins, Amsterdam, pp. 245–71
25. Wurzel W U 1989 Inflectional Morphology and Naturalness. Kluwer, Dordrecht, The Netherlands