Re-evaluating comparison between English and German : Indo-European perspectives

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.


Introduction
In this paper, English and German are compared historically along with other Indo-European languages (henceforth IE languages), focusing on their alignment changes.English and German are often compared, e.g.Hawkins (1986), but alignment is surprisingly not mentioned.This absence is perhaps because all of the modern IE languages have one type of alignment.However, historical comparison reveals that earlier IE languages had a different type, and the change from an earlier type to the modern one helps us to explain the current grammatical structure in English and German.
This paper begins with a description of different alignment types which consequently raises various issues concerning variations of transitivity.Various differences in these types are described in detail.Then, the transitivity is compared in English and German.Once this comparison is explained historically, differences between these two languages appear clearly.This paper, based on comparative analysis, also mentions possible developmental paths for the future changes, especially in German.

Alignment change in IE languages: historical perspectives
The term alignment means the way a language treats a subject and an object grammatically in terms of the distribution of morphological markers or of syntactic, semantic or morphological characteristics.Furthermore, these different systems can be roughly classified into a couple of units.The most commonly-known classifications of alignment are nominative-accusative alignment (henceforth accusative alignment) or absolutive-ergative alignment (henceforth ergative alignment).The difference between these alignment types is that the subject in transitive and intransitive constructions is treated identically in accusative alignment while the subject of intransitive constructions and the direct object of transitive constructions are identical in ergative alignment.Another alignment type, known as active-stative alignment (henceforth active alignment) is identified based on the split of intransitive subjects into two groups: the active-cum-pseudo-transitive subject and the stative/inactive-cum-transitive object.Although it may not be clear what these alignment types mean if they are explained only verbally, they can be clarified according to their relationships as schematically represented in Figure 1  Except for A (1977) and Toyota (2005), there are few studies on alignment concerning IE languages.Normally, alignment is used in studies of Caucasian and Amerind languages.The current alignment distinction as explained above is based on the one established by Sapir (1917), although he did not use the same terminology.Scholars in the former USSR, e.g.Klimov (1973), took alignment seriously, but it was not known to the rest of the world until Nichols (1992) introduced its value to the west through translation.Having claimed this, there are some earlier woks dealing with alignment, e.g.Comrie (1989), Harris (1990) and Dixon (1994), but they do not directly deal with IE languages.
All of the modern IE languages have accusative alignment.Historically, however, these languages once had active alignment.A number of reconstruction works on earlier IE languages (e.g.Gamkrelidze and Ivanov 1995;Szemerényi 1996;Lehmann 1993;1997;2002) claim that an ancestral language of modern IE languages, Proto-Indo-European (henceforth PIE), had active alignment.As far as the written records are concerned, IE languages consistently demonstrate the accusative alignment pattern, and the alignment change in IE languages is considered to be from an active to an accusative alignment.The change used to be thought to follow a pattern from active to accusative via a stage of ergative alignment known as stadial hypothesis (e.g.Kuryłowicz 1946;Klimov 1973;Schmidt 1977, 109-111;Schmalstieg 1980, 169-172), but this is discredited now.Note, however, that there is a partial ergative alignment in some IE languages known as split-ergativity.The type found in Celtic and Indo-Aryan languages is based on tense-aspect (cf.Dixon 1994: 97-101): the perfect aspect expressed in Celtic and Indo-Aryan languages has to use the undergoerorientation by default as demonstrated in (4) from Irish.The lack of a construction with the actor-orientation makes this structure superficially look like the passive voice in other IE languages, but this should be interpreted as an active construction with a different alignment system.This type of constructions often happen when historical changes leave some gaps in the verbal paradigm in a language in question (cf.Toyota and Mustafović 2006).

Semantic and syntactic transitivity
Alignment change from active to accusative is a change from an aspect-oriented grammar to a transitivity-oriented one.Active alignment organises a grammar based on perfectiveimperfective aspectual distinction.Still, a speaker can express causer-causee relationships.However, whether action has been terminated or not has priority over 'who does what to whom'.When transitivity emerged, a speaker's concern shifted to the energy transfer and the causer-causee relationship gained more prominence in grammatical organisation.Grammar of ancient recorded languages often shows an earlier sign of a causer-causee relationship.This relationship is realised in the grammatical voice as the active-middle dichotomy where the active expresses the presence of causer-causee relationship and the middle, the lack of it (cf.Gamkrelidze and Ivanov 1995).Judging by this, the alignment change must have happened beyond the recorded history of IE languages approximately 4,000-5,000 years ago.The energy transfer, i.e. transitivity, was realised earlier semantically, and later became more syntactically expressed.The semantically-oriented transitivity is principally subjective, allowing subtle differences to be expressed based on a speaker's viewpoint.The syntactically-oriented one is objective and details of personal views are not encoded in this type.The first type is termed here as semantic transitivity and the latter, syntactic transitivity.This distinction is useful in discussing historical development of IE languages since different languages are at different stages of transition from semantic to syntactic transitivity.
A characteristic of semantic transitivity is that speakers can express the transfer of energy in gradience, and there can be an intermediate stage, i.e. some sentences are more transitive than the others and some ambiguous cases can be found (cf.Hopper and Thompson 1980;Taylor 2003, 222-246).It is often the case that perception involves ambiguous cases of energy transfer, and structures involving perception are often syntactically marked.One such example is the lack of nominative subject as found in Old English, e.g. ( 5), where the NP, marked as genitive, functions as a cause and the NP in dative as a recipient of cause.This example does not contain any NP in nominative case, but it was fully grammatical earlier.Another case involves the different case marking on the direct object: the direct object marked as accusative normally refers to the action, and the object with dative expresses resulting state of action (cf.Lass 1994, 229-230 and 238).Thus (6a) places emphasise on the action of following, while (6b), on the resulting state of action.Comparing these two cases, dative denotes higher degrees of energy transfer since the transfer is considered completed.Accusative, on the other hand, refers to a lesser degree of transfer since the transfer is still ongoing and the object is not yet completely affected.These instances are largely made possible with case markings, and languages with semantic transitivity often preserve the case.Syntactic transitivity is, on the other hand, purely concerned with the presence or absence of a direct object.When it is present, a clause is transitive, and when absent, intransitive.In this type, every structure, even including perception verbs, is uniformly constructed.The grammatical subject tends to be human due to the anthropocentric nature of human language in unmarked constructions, which makes structures in English I like cakes possible.In addition, syntactic transitivity is mainly concerned with the transfer itself and the manner of transfer, i.e. whether it happens intentionally or spontaneously, is not significant.This explains why the passive voice is often found, but not the middle voice.The passive requires a high degree of energy transfer by default, since "the more transitive a clause is, the more readily it can be passivised" (Kittilä 2002, 23).However, syntactic tran-sitivity allows the passivisation of semantically intransitive verbs when the direct object is present, e.g.English This book was liked by many children from Many children liked this book.The middle voice mainly refers to spontaneous events, and is not transitive by nature, but since the manner of energy transfer is not significant, the middle voice is not often used in this type.

English and German: IE perspectives
Needless to say, both English and German have accusative alignment like other modern IE languages.However, it is obvious that there are numerous differences both structurally and functionally between these two languages in spite of the fact that they both belong to the Germanic branch of IE languages.Some features are listed in Table 1.These are selections of features which clearly show differences between these two languages.See Toyota (2008;forthcoming) for further details and examples concerning these features.What these features reveal is that varying degrees of differences can be found even in the same language family, and such differences can be explained in terms of alignment change.Toyota (forthcoming) argues that German is more archaic than English, and features found in German are characteristics of older IE languages.Compare the features of modern languages in Table 1 with the ones from older languages in Table 2.It is obvious that Old English and Old-high German share similar grammatical patterns except for the lack of the middle voice in Old English.The changes from Old-high German to Modern German are subtle: from free order with SOV basic order the word order became V-2 and the passive voice started to emerge.The degree of changes for English is much greater, and Old English is much closer to Modern German than to Modern English.The answer for such diversity within the same alignment system is the transitivity and its development.Judging from various features involved in the semantic and syntactic transitivity and features listed in Table 1 and Table 2, both Old-high German and Modern German seem to have semantic transitivity.Modern English, however, operates on the syntactic transitivity, although Old English used, and Modern German still uses, semantic transitivity.Thus, it is obvious that the main difference between English and German lies in transitivity.Historical change from semantic to syntactic transitivity, as witnessed in English, has not happened in German, and consequently, the older semantic transitivity is still found.Considering the fact that grammaticalisation is unidirectional, English has developed further in a possible developmental path for the IE languages, or in other words, German has not changed much and still preserves earlier semantic transitivity.The comparison between English and German may be an extreme case, since other Germanic languages are close to English which makes the comparison less obvious.However, a case examined here between English and German is a good example that a somewhat radical comparison is even possible within a single language family.

Recapitulationist hypothesis and future development
As far as the alignment change is concerned, all IE languages shifted into accusative alignment from active one.Although this is uniform, there are differences among IE languages which are all due to different degrees of development of transitivity.Historical changes often have regularities, and one can observe patterns.These patterns make it often possible to predict what might happen in the near future.
The early 70's and 80's saw interests on language development incorporating a biological concept of recapitulationist hypothesis (i.e.foetus recapitulates phylum, first proposed by a biologist Ernst Haeckel in 1874), e.g.Lamendella (1976), Givón (1979), Bickerton (1981;1990;1995), and more recently, Toyota (2007) and Hallonsten (forthcoming).Judging from the developmental pattern, it seems possible to claim that Modern German is still at a stage of earlier English.Furthermore, it will follow similar developmental paths and become more or less identical to PDE in terms of grammatical structures.Let us take an example concerning the progressive aspect.English has a marked periphrastic structure for this aspect, but an equivalent structure is not found in standard German.However, some colloquial form has a prepositional phrase am 'at' + INF expressing a progressive aspect, as shown in (7).In Old and Middle English, there was a structure similar to this, such as He is on hunting or its reduced form He is a-hunting (Jespersen 1949: 168;Rissanen 1999, 217;Hallonsten forthcoming), from which English developed its current form.Judging from the history of English, a structure found in (7) can be an initial stage of future progressive form in German.
Colloquial German (Heine & Kuteva 2005, 65) (7) Er ist am Essen he is at eating 'He is eating.'Recapitulationist hypothesis may not have gained much attention, but this approach is applicable to other features of grammatical changes, such as word order change.With careful consideration, it may reveal something significant in historical studies, and thereby deserves its due consideration.

Conclusion
It has been demonstrated that alignment change can provide an interesting insight in historical changes of Indo-European languages.Due to the change from active to accusative alignment, IE languages became sensitive to the causer-casuee relationship, and language structures are mainly based on how transfer of energy is encoded in grammar.This corresponds to the emergence of transitivity, which first appeared as a semantic transitivity and later developed into a syntactic one in some languages.Differences found between English and German can be attributed to the difference in types of transitivity, and English has a more advanced syntactic type, while German still operates on older semantic type.