Insa Gulzow: The Acquisition of Intensifiers. Emphatic Reflexives in English and German Child Language.

Gast, Volker

Insa Gulzow: The Acquisition of Intensifiers. Emphatic Reflexives in English and German Child Language. Studies on Language Acquisition 22. Berlin and New York: Mouton de Gruyter, 2006. xi + 279 pp.

Insa Giilzow's book documents the use of self-forms and the German intensifier selbst/selber as well as related "agent-sensitive" expressions by children up to five years old. Starting in the third year of life, intensifiers become an important linguistic device that children use to comment on the degree of agentive participation in an event, and to express their desire to act independently (no, mommy, wipe butt off self, ich selber anzieht). Given that the type of expression under consideration has been claimed to interact directly with verbal argument structure (e.g., Hole 2002), intensifiers can also be regarded as early linguistic indicators of thematic event organization. While the elements under discussion seem to play a more marginal role in adult language, they are, thus, rather important linguistic devices in early stages of language acquisition.

Studying the acquisition of intensifiers is also interesting for another reason: given that intensifiers (The president himself said it) are formally indistinguishable from reflexive anaphors (He saw himself) in English and differ from the latter only in their distribution, investigating their acquisition amounts to intruding into one of the core areas of Generative Grammar. The theory of binding has instigated a lot of empirical research aiming to substantiate Chomskyan innateness claims, and the role of Universal Grammar in language acquisition. Conversely, research into binding has become a major playground for nativist-oriented language acquisition research, as is witnessed, for instance, by the first two issues of the journal Language Acquisition (published in 1990 and 1992), which contain several articles dealing with the acquisition of the Binding Conditions (with vol. 2.4 being a special issue on the development of binding). Approaching the acquisition of self-forms from the other direction--namely, from the perspective of their use as intensifiers rather than reflexives--represents an interesting change in perspective that challenges the way the acquisition of self-forms has been modeled by Generative grammarians quite fundamentally. For instance, as is shown by G[ulzow], intensifying -self and other "agent-sensitive expressions" (esp. by -self) are acquired before reflexives. What kind of relation, then, holds between intensifiers and anaphors in early language acquisition? Are anaphors acquired on the basis of non-anaphoric uses of self-forms? Or are they acquired independently? Are there different types of developments across languages, depending on whether or not intensifiers and anaphors are formally identical, etc.?

Before far-reaching questions like these can be addressed in any detail, it is, of course, necessary to determine the way intensifying and reflexive self-forms are actually acquired, and this is precisely what G aims at. Her book is written in a largely descriptive and documentary spirit. Unlike most other studies dealing with the acquisition of self-forms, it is based on production data, which has been extracted from the CHILDES database. The data are presented in a theory-neutral way (though the author is committed to a non-modular view of language acquisition), with the usual advantages and disadvantages of such an approach: on the positive side, the unbiased exhibition of the data renders them suitable as a possible input to future studies; on the negative side, the author's agnosticism seems to prevent her from pursuing more thoroughgoing questions such as those mentioned above in very much detail. The documentary spirit of the book and its theory-neutrality add to its character as an exploratory study which raises new questions rather than answering old ones.

After a short overview of the book Chapter 1 starts with providing some basic information about intensifiers. Their morpho-syntactic properties in different languages are outlined and, more importantly in the present context, their semantic properties are described. G bases her analysis on joint work by Ekkehard Konig and Peter Siemund (referring to Siemund 2000 and Konig 2001), disregarding some more recent work such as Eckardt (2001), Hole (2002) and Gast (2006). Konig and Siemund distinguish two major use types of intensifying self-forms, namely (i) an adnominal use ([NP The president himself]) and (ii) an adverbial use, where the intensifier forms part of a verbal projection. The latter use is moreover sub-classified into an "exclusive use" (~by self, e.g., He [did it himself]) and an "inclusive" one (~too, e.g., I [have children myself]). G focuses on the "adverbial exclusive" use of intensifiers (sometimes also called "agentive", e.g., by Kemmer 1995 and Hole 2002), since it is the use figuring most prominently in early (English and German) child language. Exclusive intensifiers are subcategorized further into two semantic subclasses, namely (i) "anti-assistive" and (ii) "autonomous" uses. In the first case, (exclusive) self expresses that some referent x, rather than a plural referent (x+y), carries out an action (let me do it alone/without your help); in the second case, it expresses that x carries out the action instead of some alternative agent y (let me do it instead of you). Both of these uses are particularly central to child language, as they are important linguistic means for children to gain independence, and to start carrying out activities their parents are accustomed to doing for them.

In addition to "bare" intensifiers (Engl. -self and the Germ. synonyms selbst/selber), G discusses the semantic and pragmatic properties of what she calls "agent-sensitive" expressions, for instance Germ. allein and yon selbst, and Engl. by -self and on one's own. Some subtle semantic differences are pointed out, both between singular items of each language (e.g., Germ. selbst vs. von selbst), and between comparable items of English and German (e.g., Germ. allein vs. Engl. by -self). While the first part of the introduction, where the semantic and syntactic properties of intensifiers in general are summarized (Sections 2 and 3 of Chapter 1), is rather shallow and also a bit dated, the description of agent-sensitive expressions in Section 4 is quite informative and contributes some new interesting observations. Different readings of specific expressions are classified according to properties of the associated predicates (esp. matters relating to argument structure and aktionsart), a line of reasoning also pursued by Siemund (2000). For instance, by -self is shown to be usually "predicative" in combination with stative predicates (roughly, meaning "alone", e.g., She is by herself), while it is typically "exclusive" in combination with telic predicates (~himself, e.g., He dug up the garden by himself). In addition to being indispensable for an understanding of the study itself, this part is also interesting from a purely linguistic perspective, since it broadens the picture in comparison to earlier relevant studies (e.g., Siemund 2000; Hole 2002; Gast 2006), which basically focus on the "core members" within the class of intensifiers mentioned above (-self, selbst/ selber).

Following the make-up of a dissertation, Chapter 2 summarizes previous research on -self in acquisition studies. As is to be expected, most previous research has been carried out within the framework of Generative Grammar and deals with either reflexive or logophoric uses of self-forms. The most important pertinent studies discussed by G are based on comprehension data, which is certainly one reason why the role of intensifying self-forms has often so severely been underestimated: given that intensifiers are not represented in the stimuli, they do not figure prominently in the data. G restricts herself to summarizing the arguments made by the various authors without taking a stance concerning the methodologies applied, or the results achieved. She does point out, however, that even analyses based on the innateness hypothesis attribute considerable importance to pragmatic principles in the process of acquisition, i.e., a simple model of (syntactic) parameter-activation is not supported by the data. This becomes relevant later, in particular, in Section 6 of Chapter 3, where G relates her own (production based) results to those obtained in comprehension studies.

After the summary of previous research (a bit late perhaps, on p. 88), G states a set of five research questions guiding her investigation, thus orienting the reader for the discussion to follow. These questions concern: (i) the use types of intensifiers observed in the data of both languages under consideration (adnominal, inclusive, exclusive), (ii) the relationship between intensifiers (self, selbst/selber) and other agent-sensitive expressions (by -self, allein), (iii) relevant developments observed in the data of individual children, (iv) the relationship between intensifying and reflexive self-forms in the English data, and (v) the use of non-bound bare self-forms in an argument position (it will hurt yourself). These questions are certainly well chosen and cover the most important aspects of the domain under investigation. Moreover, the author is to be congratulated for stating a set of explicit research questions in the first place, which greatly facilitates the reading of the (very detailed and therefore somewhat lengthy) presentation of the data in Chapter 3.

Chapter 3 constitutes the core of the study. This chapter is characterized by one major strength and one potential weakness: the strength consists in the fact that the data are not pooled across children, as is common practice in both psychology and (corpus) linguistics, i.e., each child is analyzed individually. This procedure allows for the observation of individual developments as well as of differences between children of the same language. It also provides a nice view of the way children form their own grammars "by hypothesis", i.e., by creating transitional idiolects with specific rules and patterns that are later given up. The potential weakness concerns the treatment of any type of quantitative data. Given the scarcity of (production) data available, the study is necessarily exploratory and all observations are somewhat tentative, since statistically significant generalizations are hardly possible. Nevertheless, G yields to the temptation of making a number of quantitative statements, without however providing an adequate statistical interpretation (as a matter of fact, G announces a "quantitative analysis" on p. 3). In my opinion, G would have been well advised to resist that temptation, sticking to a purely qualitative agenda, since the very crude statistics (G does not even provide relative frequencies for most of the data) may be regarded as a serious shortcoming by some readers.

For each language, data from six L1-subjects have been analyzed, though for three of the German children hardly any data are available, and the total number of English utterances in the corpus is more than twice as high as the number of German utterances (approx. 180,000 vs. 85,000). The data have been coded according to two major semantic parameters: (i) the type of reading instantiated by each item (intensifier, reflexive, other), and (ii) the type of aktionsart of the relevant predicate (telic, atelic, stative). On this basis, the five research questions mentioned above are addressed for each child separately (the three German poor-data children are lumped together). As for the type of intensifier used most frequently, English and German deliver basically the same results: in each case it is the exclusive intensifier which figures most prominently in the data. Adnominal uses, arguably the most "typical" ones in adult language, are not attested at all, and inclusive intensifiers are found only in very few instances, where they are moreover associated with "control", thus being not too different from exclusive intensifiers. These findings are certainly not unexpected, given the communicative needs of children at that age, but it is nevertheless valuable to have them confirmed and illustrated with concrete data.

The second research question concerns the relationship between intensifiers and other agent-sensitive expressions, against the background of aktionsart-related information in the sentential context. One major finding of this section is that intensifiers (selber, -self) seem to be acquired later than the "agent-sensitive expressions" alleine and by -self. While English and German are shown to behave similarly in this respect, a quantitative difference is pointed out by G with respect to the use of intensifiers: bare intensifiers appear to be rarer in English than in German. One may wonder if this tendency can be related to the absence of any morphology in the German intensifiers, which makes them look similar to adverbs. It should be borne in mind, however, that all quantitative statements made in this book have to be taken cure grano salis, and should be regarded as hypotheses to be tested in future research, rather than definite results.

Probably the most important part of the book--the detailed presentation of the data--starts in Section 4 of Chapter 3 (Section 5 of this chapter provides an overview). In addition to the parameters of classification established earlier (the type of intensifiers involved and aktionsart-related properties of the predicates), another aspect of description becomes relevant here, namely the communicative intention or speech act associated with the relevant utterances. In particular, G makes a difference between the acts of negotiating and commenting on agentive involvement in an event. This contrast can also be phrased as one between "uncontroversial" and "controversial" (verbal) interaction: in the first case, there is no dissent between children and parents, for instance if a child proudly comments on a developmental achievement she has made, for which she is also praised by her parents. In the second case, children demand a more active participation in an event than their parents would be willing to grant. As G shows, this difference seems to be reflected in the data in so far as certain expressions tend to be associated with "uncontroversial" interaction while others are used more frequently in "controversial" discourse. Note that such differences are often not of a general nature but concern only individual children, and can typically be observed only at a specific stage of development.

The type of observation made in this part of the study can be illustrated with three examples: (i) one of the German children seems to make a systematic distinction between (al)lein 'alone' and ganz allein 'all alone': while allein is used for "controversial discourse" (leine--och das ist aber nichts), ganz allein is used when the child merely comments on a developmental achievement, typically in agreement with the parent (jawoll--ganz alleine kann ich--ja wirklich toll). (ii) One of the German children makes a functional difference between allein and selber, using the former for exclusive meanings and the latter in the function of an inclusive intensifier, a differentiation which is not found in adult language. (iii) One of the English children makes excessive use of by itself, in comparison to the other children. G hypothesizes that this may reflect a personal interest entertained by the child in question for matters of cause and effect, a point plausibly illustrated with relevant examples. On a more general level, some generalizations across languages also seem to be possible: in general, the German data contain more "controversial" uses of intensifiers and agent-sensitive expressions than the English data. As G points out, it is unclear whether this is due to a linguistic difference, to a cultural difference or--most likely perhaps--to experimental conditions.

The last two research questions addressed by G concern the relationship between (formally identical) reflexives and intensifiers, and the use of bare self-forms in a (non-bound) argument function. G claims that English reflexives are acquired at a relatively late stage, in comparison to bare intensifiers and by -self. Given the absence of any solid statistical evidence, this claim, again, remains a bit impressionistic, but the tendency seems to be rather stable across the children investigated here. G relates her findings to those of Chien and Wexler's (1990) and McDaniel et al.'s (1990) comprehension studies, arguing that the violations of specific binding conditions observed in these studies can be explained on the hypothesis that children first acquire the pragmatic properties associated with -self, which are then "translated" into syntactic pnnclples. Even tlaougla this hypothesis is not elaborated very explicitly, it sounds rather plausible. Roughly, G suggests that children first associate self-forms with the notion of "centrality" of a referent in discourse (a notion figuring centrally in the work done by Konig and Siemund). Later, the pragmatic notion of "centrality" is "syntacticized", leading to the (near) complementary distribution of intensifiers, (free) pronouns and reflexive anaphors. Note that such a development would correspond rather closely to the historical development of intensifiers into reflexives as described by Konig and Siemund (1999) (though G explicitly denies a link between ontogenesis and phylogenesis on p. 122).

Before coming to a survey and final evaluation of the book in terms of content, a word is in order about its poor editorial quality. The book exhibits editorial imperfections of all types. There are plenty of typos (e.g., descibed on p. 7, possibilitiy on p. 12, substancial on p. 31, book fell of on p. 34, appropiately on p. 44, and and on p. 45, to name just a few), inconsistencies in orthography (willfullness on p. 31 vs. willfulness on p. 32), punctuation (e.g., in the order of closing quotation marks and full stops in examples) and cross-referencing (e.g., Siemund [2000] on p. 4 and Siemund [1999] on p. 13, referring to the same book, Table 39 instead of Table 40 on p. 184, etc.). Another annoying feature is the entire loss of one set of fonts, which leads to complete unintelligibility in some examples from Old English (e.g., xFt instead of doet and Selx Goxe his Fhta ... instead of Seld Gode his oehta ... on p. 83). The book would also have profited from a careful checking of grammar and style. These editorial imperfections should, in my opinion, not be blamed on the author alone. We all make mistakes of different types, and most manuscripts sent to publishers contain precisely the type of imperfections listed above. However, economizing efforts (and maybe sheer haste) seem to have reached a level where such shortcomings are no longer corrected on the publishers' side, so that quality is seriously compromised.

Coming back to matters of content, I would like to conclude on a more positive note. This book is an exploratory study of a very complex topic, the first to tackle the acquisition of self-forms and related expressions in English and German to this level of detail. Unlike most other pertinent studies, it is based on production data. While this implies a certain scarcity of data and, consequently, the absence of solid quantitative evidence, it raises a number of interesting new questions and leads to the formulation of stimulating new hypotheses. It is to be hoped that this study will lead to further experimental research.

VOLKER GAST

Free University, Berlin

References

Chien, Y.-C. and Wexler, Kenneth (1990). Children's knowledge of locality conditions in binding as evidence for the modularity of syntax and pragmatics. Language Acquisition 1(3), 223-295.

Eckardt, Regine (2001). Reanalysing selbst. Natural Language Semantics 9, 371-412.

Gast, Volker (2006). The Grammar of Identity. Intensifiers and Reflexives in Germanic Languages. London: Routledge.

Hole, Daniel (2002). Agentive selbst in German. In Sinn und Bedeutung VI, Proceedings of the Sixth Meeting of the Gesellschaft fur Semantik, Osnabruck, Germany, October, 2001, G. Katz, S. Reinhard, and P. Reuter (eds.), 133-150. Osnabruck: University of Osnabruck.

Kemmer, Suzanne (1995). Emphatic and reflexive -self. Expectations, viewpoint and subjectivity. In Subjectivity and Subjectivization in Language, Dieter Stein and Susan Wright (eds.), 55-82. Cambridge: Cambridge University Press.

Konig, Ekkehard and Peter, Siemund (1999). The development of complex reflexives and intensifiers in English. Diachronica XVII, 39-84.

Konig, Ekkehard (2001). Intensifiers and reflexive pronons. In Language Typology and Language Universals, Martin Haspelmath, Ekkehard Konig, Wulf Oesterreicher and Wolfgang Raible (eds.), 747-760. Berlin and New York: Mouton de Gruyter.

McDaniel, Dana; Cairns, Helen Smith; and Hsu, J. R. (1990). Binding Principles in the grammars of young children. Language Acquisition 1(1), 121-139.

Siemund, Peter (2000). Intensifiers: A Comparison of English and German. London: Routledge.