Review of Actualization. Linguistic Change in Progress.
Date: Fri, 19 Apr 2002 14:04:58 -0400
From: Marc Picard
Subject: Andersen (2001) Actualization: Linguistic Change in Progress
Marc Picard, Concordia University
This book is the outgrowth of a workshop on "Patterns of Actualization in Linguistic Change" that was held in conjunction with the Fourteenth International Conference on Historical Linguistics in Vancouver, BC in 1999. It consists of a collection of ten papers whose general aim is "to consolidate the observation that the progression of certain kinds of linguistic change is grammatically conditioned" (p. 1). This is based on the premise that "there is some correlation between the environments in which innovations occur earlier or later and the markedness values of the phonological, phonotactic, morphophonemic, morphosyntactic, clause-syntactic, lexical-semantic, referential, and other parameters involved" (p. 3). The articles are preceded by the editor's introduction and followed by a short general index.
Andersen's "Markedness and the theory of linguistic change" sets the
stage by introducing the Principle of Markedness Agreement which he
defines as "the favoring of combinations or concatenations of different
features that are homogeneous in Markedness value" (p. 49). He begins by
showing how this concept is manifested synchronically in such diverse
areas as ritual, the thematic and plot structures of texts, the
grounding structure of narrative discourse, and the regularities of
morphosyntax, morphophonemics and phonology. Then he turns to markedness
in diachrony and argues that "as a linguistic innovation gains currency
and is generalized in a language, the process of actualization conforms
to the Principle of Markedness Agreement in that the innovated element
is favored first of all in marked environments, if the innovated element
is marked, but in unmarked environments if it is unmarked" (p. 31). In
sum, then, Andersen sees markedness as a principle of cognitive
organization which is fundamental to human behavior and which manifests
itself linguistically not only as a synchronic property of speakers'
grammars but as a significant conditioning element in language change.
The second article is Kristin Bakken's "Patterns of restitution of sound change" wherein she examines the partial reversal of two historical phonological processes from the dialect area of Western Telemark in Southern Norway. The two sound changes in question are the loss of postvocalic /l/ before consonants other than alveolar stops, e.g., Old Norse (ON) holmi 'island' > /hu:m@/, and the shift of postvocalic /ll/ to /dd/ (probably via /dl/), e.g., ON allir 'all' > /add@/ (@ = schwa). What she succeeds in showing is that the undoing of the effects of these processes, that is, the return to the etymologically older forms in certain words, "most likely was a lexically diffused substitution process triggered by dialect contact and linguistic variability [that was] obviously bound to be sociologically and culturally determined" (p. 75).
Next comes "The role of markedness in the actuation and actualization of linguistic change" by Alexander T. Bergs and Dieter Stein. This consists mainly in an investigation of the origin and early history of the English relative markers *which*, *who*, *whose* and *whom* which gradually replaced the Old English relativizers *the* and *se* at different times beginning around the 13th century. For the authors, "[t]his development is interesting for a theory of actuation and actualization insofar as it shows that new forms can originate (be actuated) in ... salient, marked contexts, and ... be gradually actualized in the linguistic system as their markedness declines" (p. 82). This is particularly evident in the case of *who* which was first triggered by God, saints and other religious antecedents, then by noblemen, good friends and worthies, and finally by any [+human].
This is followed by Vit Bubenik's "On the actualization of the passive-to-ergative shift in pre-Islamic India" in which he examines the proposal that ergative systems can arise through a shift in markedness that operates on the active-passive opposition of a nominative language. Basing his analysis on a corpus of fourteen Prakrit poetic texts written between the 7th and 11th centuries, he attempts to show that the actualization of the passive-to-ergative shift occurred more widely and earlier:
(1) with agents high in animacy;
(2) in agent-initial sentences;
(3) with discourse participants;
(4) in the interrogative mode;
(5) in causative predicates.
Correspondingly, this shift was more limited and later:
(1') with agents low in animacy;
(2') in agent-final sentences;
(3') in the third person;
(4') in the declarative mode;
(5') in noncausative predicates.
His conclusion is that these changes "are in keeping with Andersen's Principle of Markedness Agreement ... which predicts innovations will spread earliest in environments with equivalent markedness value and subsequently spread to the complementary environments with opposite markedness value" (p. 111).
"The use of address pronouns in Shakespeare's plays and sonnets" by Ulrich Busse analyzes the alternation between *thou* and *you* in the Early Modern English period as evidenced mainly in Shakespeare's various writings (1589-1613). Although these pronouns were both available to speakers and writers of the time, they were not necessarily interchangeable socially, stylistically or rhetorically. Thus, "by making a link between a statistically more or less probable form (*thou*) and its stylistic value as the marked item in the dyad", the author arrives at "the following text typology in terms of *thou*-fulness" (p. 138):
(1) *thou* is marginal in nonfictional writings from 1500 on;
(2) in Shakespearian drama, the History Plays and the Tragedies are the most *thou*-ful, while the Comedies are more *you*-ful;
(3) in terms of their distribution in the media of verse and prose (in drama), *thou* predominates in verse and *you* in prose;
(4) in the elevated register of poetry as compared to drama, *thou* is more prevalent;
(5) more overtly public poems prefer *you*, more private ones, *thou*.
Marianne Mithun's article "Actualization patterns in grammaticalization: From clause to locative morphology in Northern Iroquoian" examines the emergence and development of a new locative category in Mohawk, though the same types of changes apparently occur in other Iroquoian languages. The locative markers in question, such as *-ke/-hne* 'place', *-kon* 'inside', *-okon* 'underneath', *-akta* 'near, beside', *-(i)ken* 'middle', *-ti* 'beyond', have evolved at a different rate in accordance with their degree of markedness. Thus, for example, the most grammaticalized marker, the one for 'place' (*-ke/-hne*), is also the most general in meaning, the most frequent in occurrence, and the only one to show significant allomorphy, a feature that has been cited as characteristic of unmarked elements.
The gist of Lene Schøsler's "From Latin to Modern French: Actualization and markedness" is that a nominal markedness hierarchy proposed by Andersen in the article outlined above is partly violated by the evolution of the Latin noun case system in French. The hierarchy in question is as follows:
(a) proper common
(b) human non-human
(c) animate inanimate
(d) concrete abstract
(e) singular plural
(f) definite indefinite
and this seems to create a dilemma. On the one hand, categories like plural and non-human lose case first which might indicate that in Old French the hierarchy was reversed given that "Andersen suggests that hierarchies may be language specific". On the other hand, however, "even then proper nouns remain a problem, because proper nouns lose case first instead of last" (p. 174). In commenting on this problem, Andersen notes that "quite unlike common nouns, names have no descriptive content [which] suggests the possibility that [they] may be categorized (in some languages or universally) as a subclass of pronouns - in which case their role in the hierarchy of categories that condition the progression of case loss in Old French is in accord with the Principle of Markedness Agreement" (p. 14).
"Markedness, causation, and linguistic change: A semiotic perspective" opens with a reflection on how differently philosophers and linguists generally approach the study of language, in response to which Michael Shapiro states his goal as follows: "Perhaps a bridging of the gap between the two disciplines can be essayed here by way of approaching the several points I would like to make about markedness, causation, and linguistic change" (p. 189). His paper is divided into four sections, the first two of which - a philosopher's view of language, and nominalism and realism in linguistics - are only peripherally related to the issues at hand. While the other two do address the topics mentioned in the title, the discussion remains philosophically oriented and no actual language data are introduced to illustrate his point that "markedness must be viewed as a final cause in linguistic change" (p. 199).
In "Markedness, functionality, and perseveration in the actualization of a morphosyntactic change", John Charles Smith attempts to assess the influence of these factors on the progressive disappearance of agreement between a past participle and a direct object in the Romance compound past tenses formed with the auxiliary *have*. For example, synchronic and diachronic data from Catalan show that the plural past participle forms *vistes*/*vists* in sentences like:
(a) He vistes les pellicules 'I've seen the films'
(b) Quines pellicules he vistes? 'What films have I seen?'
(c) Les pellicules que he vistes 'The films I've seen'
(d) Us he vistes 'I've seen you (fem. plur.)'
(e) Els he vists 'I've seen them (masc.)'
(f) Les he vistes 'I've seen them (fem.)'
have gradually been reduced to singular *vist* in accordance with a set of implicational hierarchies having to do with:
(1) the position of the direct object;
(2) the identity of the preceding direct object;
(3) the person of the clitic pronoun;
(4) the number and gender of the third-person nonreflexive clitic pronoun.
The end result is that hierarchies (2)-(4) fail to account for the differential disappearance of agreement in terms of markedness if the latter is identified with frequency (a connection Andersen argues against in his "Introduction" (pp. 1-20) wherein he also addresses the troublesome aspects of Smith's analysis). Instead, Smith proposes that "this particular instance of actualization appears to be sensitive to functionality, rather than markedness, in respect of morphosyntactic environment" (p. 217), the basic premise of functional change being that "the need to preserve information is an influence on how language develops" (p. 206).
The closing article is Andersen's "Actualization and the (uni)directionality of change" which he himself summarizes concisely in his "Introduction": "[This] paper . . . describes the place of actualization - the only observable aspect of change - in a theory of change; shows how the theory of Markedness proposed in [the first article in this volume] makes it possible to understand change as a projection of synchronic variation onto the diachronic axis; and tries to clarify the relation between historical change events, the domain of the language historian, and the generalized 'change schemas' that the historical linguist can use to advantage in investigating the origins of types of linguistic change" (p. 10).
This is not an easy read by any means, even for a seasoned historical linguist. One reason is the range of complex and disparate data one has to contend with for it is no easy matter to critically evaluate the validity of various syntactic and morphological markedness hierarchies in such diverse language groups and families as Iroquoian (Mohawk), Germanic (Norwegian, Old English, Early Modern English), Indo-Aryan (Prakrit), Italic (Latin), and Romance (Catalan, Old French). Another reason this off-the-beaten-track material may be fraught with difficulty for many a linguist is that markedness is such a tenuous and elusive concept, one that is not easily amenable to neat, consistent and straightforward analyses. There is simply no easy, once-and-for-all way of determining whether such and such a phenomenon is marked or unmarked.
In phonology, for example, no one would dispute the fact that voiceless stops are unmarked in relation to their voiced counterparts (many languages have only voiceless stops, no languages have only voiced stops). Intervocalically, however, voiced stops are the unmarked variants (languages with both series may have only voiced stops but not only voiceless stops in that environment). In syntax, there seems to be a general agreement that main clauses are unmarked in relation to subordinate clauses since changes seemingly occur there first, e.g., SOV > SVO in many Germanic languages. However, as reported by Smith in this volume (p. 205), the merger of attributive and predicative forms in Old Japanese occurs in dependent clauses before main clauses. Consequent on this, he observes that "there is an abiding problem with any analysis based on markedness: the definition of the concept is not unproblematic, and in many cases (although not all) the choice of one or other member of an opposition as the marked term can appear arbitrary" (p. 205). Still and all, attempts to provide principled and systematic accounts of the actualization of linguistic change, whether it be in terms of markedness, functionality or some other governing principle, should continue to be made, and Andersen's extensive investigation of the subject will assuredly serve as a catalyst for the further exploration of these important and fascinating issues.
| ABOUT THE REVIEWER:
Marc Picard teaches phonetics, phonology and general linguistics at Concordia University in Montreal. He is the author of "Principles and Methods in Historical Phonology: From Proto-Algonkian to Arapaho" (McGill-Queen's University Press, 1994) and has also published a number of articles on historical phonology and morphology in various North American and European linguistics journals. The most recent are "On the fricativization of /r/ and the French-Cree connection" in Folia Linguistica Historica 22:137-147 (2001) and "Aspects synchroniques et diachroniques du hiatus: le cas du déterminant /la/ en créole haïtien" which will appear in an upcoming issue of La Revue québécoise de linguistique.