A Comparative Typology of English and German

Unifying the Contrasts

John Hawkins

• Publisher: Routledge
• Year: 2015
• ISBN:
  9781317419716
• Edition: 1
• Topic: Lingue e linguistica
• Subtopic: Linguistica

First published in 1986, this book draws together analyses of English and German. It defines the contrasts and similarities between the two languages and, in particular, looks at the question of whether contrasts in one area of the grammar is systematically related to contrasts in another, and whether there is any 'directionality' or unity to contrast throughout grammar as a whole. It is suggested that there is, and that English and German can serve as a case study for a more general typology of languages than we now have.

This volume will be of interest to a wide range of linguists, including students of Germanic languages; language typologists; generative grammarians attempting to 'fix the parameters' on language variation;' historical linguists; and applied linguists.

PART ONE

AN OVERVIEW OF SOME ENGLISH/GERMAN CONTRASTS

1 INTRODUCTION: THE THEORETICAL INTEREST OF ENGLISH/GERMAN CONTRASTS

1.1 English/German Contrasts in Relation to Universal Grammar and Language Typology

Much recent research in grammatical theory has shared a common goal, despite numerous differences in approach. It is the goal of building a theory of Universal Grammar, i.e. a theory defining the properties and regularities which hold for all languages. The pursuit of this goal in the modern era can be traced back to Chomsky’s (1965) stipulation that a linguistic theory should define the notion ‘possible human language’. For Chomsky this amounted to specifying the basic building blocks of the grammars of all languages in terms of ‘formal’ and ‘substantive’ universals. Formal universals are those that relate to the form and shape of the grammar, e.g. the various components of the grammar (syntactic, semantic, etc.), the different rule types (phrase-structure rules, transformations), principles of rule interaction (the cycle), etc. Substantive universals refer to the contents of these rules, syntactic categories such as NP, distinctive features within the phonological component, etc. We will refer to these two types of universals as ‘absolute universals’. Also included under this label will be less abstract universal generalisations such as ‘all languages employ vowels in their phonological inventories’ and ‘all languages have pronominal categories involving at least three persons and two numbers’ (cf. Greenberg 1966:113). All of these universals define grammatical properties that are exemplified in every single language.

The methodology that Chomsky and his associates employed in the 1960s and 1970s in discovering such absolute universals involved an in-depth grammatical analysis of a small number of languages, typically just English alone. Generative grammarians were concerned to build a ‘descriptively adequate’ grammar of English, one which accounted for all and only the relevant phenomena within this language (all and only the grammatical sentences, etc.), and which incorporated the formal and substantive universals of the general theory. Progressive refinements in this task led to changing hypotheses about these formal and substantive universals. The rationale behind this approach was that since universals are properties that all languages share, they can be derived from the study of any one language. And by defining these universals as precisely as possible, the theory distinguishes possible human languages (whose grammars incorporate the universals) from impossible human languages (whose grammars do not).

As a reaction against this methodology, numerous linguists argued for a more comparative approach. They pointed out that there is considerable variation between languages, yet there appear to be profound regularities underlying, and constraints on, this variation. These constraints can also be regarded as universals of language, since they delimit the types of grammatical properties and property combinations that languages can exhibit and so contribute to a definition of the notion ‘possible human language’. A possible human language in this view is one whose properties fall within the parameters permitted by the variation-defining language universals, an impossible language is one whose properties do not. To quote from Keenan (1978:90): ‘Universals are characterisations of the regularities in the ways languages may differ from one another. Structures which differ from one another … are among the primary objects of study.’ Classifications of languages made in terms of the properties or property clusters mentioned in these variation-defining universals are then referred to as ‘typologies’. This paradigm of universal grammar is often called ‘Typological Universal Grammar’. Other individuals whose names are associated with it include Greenberg, Comrie, Lehmann, Vennemann and Thompson.¹ For a discussion of the relationship between typologies and universals, cf. Hawkins (1983:45–47; 50–51).

Some paradigm examples of proposed variation-defining universals are the implicational universals of Greenberg (1966), e.g. ‘if a language has verb-subject-object word order (VSO), then it has prepositions (before NP rather than postpositions after NP)’. More recent versions of generative grammar (cf. Chomsky 1981, 1982) are now also concerned to define possible and impossible language variants by setting ‘parameters of universal grammar’. These parameters are typically more abstract than those of the typological paradigm, cf. e.g. the configurationality parameter (Hale 1981, 1982, 1983).

In Hawkins (1983) I proposed a number of typological universals of language in the area of word order. Building on the language samples and universals of Greenberg (1966), various implicational universals were formulated, together with some more explanatory principles which in their interaction predict the attested versus non-attested combinations of word orders that constitute a whole language type. The most striking fact about these cross-language word orders is the discrepancy between the enormous number of mathematically possible combinations of word orders in the different phrasal categories that languages could conceivably exhibit, compared with the exceedingly limited number of word order types that they actually do exhibit. Time and time again, languages that are genetically quite diverse combine the same word order patterns or else vary in predictable and principled ways. It is such regularities and the principles which can be shown to underlie them that constitute the raison d’être for including variation statements within the set of universals of language and within the domain of a theory that attempts to define the notion ‘possible human language’. What is especially revealing about word order is the fact that the set of mathematical combinational possibilities is easily quantified and is so much larger than the set of attested word order types.

At the same time this study, and others like it, is probably missing important universal generalisations. It involves examination of a small number of variant linguistic properties within large numbers of languages. The properties may be selected word orders, selected morpheme orders within words (cf. Hawkins and Gilligan in prep.), relative clauses (cf. Keenan and Comrie 1977), subjecthood properties (cf. Keenan 1976), and so on. In each case, small pieces of language are plucked out from the overall grammar that contains them, and the range of attested variation is described, and universal generalisations, or truths, are proposed that are compatible with all and only the observed patterns. Obviously, the more such pieces of language we study, the more universal generalisations we gain. But it is not clear that we are making much progress towards understanding how the variants that an individual language selects in one area of grammar are determined by, or determine, the variants that it selects in another.

The present book offers a different, and complementary, approach to the study of variation. What motivates it is still the search for principles that underlie cross-language variation. But it adopts a methodology which is in many ways the exact inverse of the comparative-universal approach. Whereas this latter examines a small number of variant linguistic properties in a large number of languages, the present approach looks at a large number of variant linguistic properties in a small number of languages.² Yet this approach still has close affinities to both generative and typological paradigms. It refers to grammatical details of English and German of the type discussed in the generative literature and contains extensive references to generative proposals for describing the phenomena involved. And it draws on relevant typological literature for the grammatical properties involved, and is concerned to characterise the typological differences between English and German. The chief novelty of our approach lies first in its attempt to consider two whole languages from a typological-universal point of view; and second in the search for unifying generalisations that underlie the variation between the major portions of two whole languages. Clearly, this kind of approach can only be pursued for the better-studied languages of the world.

The rationale for the present study stems from my belief that the patterns of variation between English and German are neither idiosyncratic to these two languages, nor accidental, and can contribute to current hypotheses about language variation in original and suggestive ways. In particular, the contrasts between English and German are theoretically interesting for three reasons.

First, it is not commonly appreciated how precise they are. There are exact, or nearly exact, proper subset relations between the structures of the two languages in each area of major contrast to be considered. I.e. it is regularly possible to set up generalisations of the kind ‘wherever language A has rule or structure X, so does language B, but not vice versa’. Most of these proper subsets are the result of English having historically expanded or diminished structural possibilities which in German have either remained constant, or have changed to a lesser extent.

Second, despite the relatively recent time depth that separates English and German from their common West Germanic ancestor (cf. Hawkins in press), the contrasts between them do not involve small changes in limited grammatical areas, but profound readjustments across all the major areas of the grammar: the morphology, word order, grammatical relations, raising rules, extractions, deletions and rules of semantic interpretation have all been affected in important ways.

Third, I shall present evidence that there is a descriptive generalisation which unites these major areas of contrast. And the primary purpose of this book is to present this generalisation, together with the facts that support it. Notice at the outset that the existence of any such generalisation uniting variation patterns in many different areas of the grammar suggests that there may be interesting interconnections of a rather abstract sort between the different areas of the grammar. But so far the full extent of these interconnections is not being captured or predicted by current principles of variation, either typological or generative.

For example, the typology drawn by Thompson (1978) betwe...