Reviews

136 Links & Letters 1, 1994 Reviews CHNSTOPHER J. HALL 1992 Morthat the morphological head (the afix) pholog, and ~i~d. A unified appears in the same position as the syntactic one. Hall rejects the HOP as he Approach Explanation in Lindoes not accept that affixes are always guistics. London: Ro '-ltledge. heads in morphology; he actually exa- 1992. xx, 224 pages. mines the notion "head of a word and decides (on the basis of inflection and This book deals with a well-attested emsome cases of derivation) that the catepirical tendency in natural lan uages gory AFF does not have any inherent which is the universal preference B or sufproperty of headship. Nevertheless, the fixation over prefixation. This interesting correlation affixlsyntactic head remains cross-linguistic phenomenon is andHallseeks to explainit. In thissense, approached from a multidisciplinary he accepts Givón's idea that affures arise and psycholinguistic perspective, and it historically due to the semantic and phomay be in a way regarded as an excuse to nological decay of syntactic heads i.e., develop and justify an integrated, nonthey come from free lexical items in head restrictive view of explanation in linguistics. position. What Hall does not accept is This review is organised as follows. Givón's explanation for the suffixing pre- First, the relevant em irical data are briefference i.e., the idea that al1 languages ly described and H J 1's interpretation of exhibit head-finality currently or at some the data is presented. Then, the author's past stage, so that affures arise at that conception of the controversial issue of moment and as a result there is a predoexplanation in linguistics is advanced, so minance of suffures. As Hall remarks, there that the general explanatory framework is very little evidence in favour of this in which the data will be embedded is claim, for example, the current ratio of made clear. Finally, the main guiding line head-final to head-initial languages is in Hall's account -the idea that the human roughly fifty-fifty. mechanism of language rocessing plays B The tendency towards suffuration ives an important role in the etermination of Hall a good chance to develop a w ole language structureis developed. B, theory of explanation. He distinguishes The universal preference for sufixatwo approaches to the explanation of tion is represented by data such as those structural regularities: a formal, linguisprovided by Hawkins and Gilligan tic-internal approach (mostly adopted in (1 988). Hall captures this tendency by generative linguistics) and a functional, means of a Suffuring Principle, according interactive one (adopted before Chomsky to which "Bound morphemes are added and in general linguistics). The two con- to the ends rather than the beginnings of ceptions are examined and a unified powords, with greater than chance frequensition which emphasises the role of cy" (Hall 1992:48). A related observation function is defended. Within the formal in Hawkins and Gilligan is that exclusive approach, regularities are explained when sufixing is predominant in head-final they follow from formal universal conslanguages and exclusive prefixing occurs traints on the grammar. One important only in head-initial languages (note that shortcoming of this conception is that it the latter fact, however, does not interfere is too restrictive in that it tends to displawith thegeneral tendencyforsufixation, ce, rather than complement, other since, according to the authors, exclusive approaches. suffixing is considerably more frequent In the functional approach, on the and besides, many head-initial languaother hand, explanation is provided by a ges are both prefixing and suffixing). number of parameters, of which the ma- Hawkins and Gilligan's observation leads jor ones are either psychological or otherthem to the postulation of a Head Orderwise functional. Whereas the generative ing Principle (HOP) which determines tradition claims to be explanatory within Reviews Links & Letters 1, 1994 137 the terms of the descriptive theory em- On the whole, this book deserves the ployed, hnctionalists often appeal to psycareful attention of al1 those whose interests cholinguistic mechanisms and processes include morphology and mind. It offers as an important source of constraints on an interesting and contentful psycholinlanguage structure i.e., explanation is exguistic program for addressing morphoternal to the linguistic theory logical problems. Al1 in all, in his endeavour to adopt a unified position, Hall retains the descriptive dimension of the formal approach References but claims that psycholinguistic principles of lexical access and organization ul- HAWKINS, J. A. and G. GILLIGAN (1988). timately explain the preponderance of "Left-right asymmetries in morpholosuffues over prefixes in the languages of gical universals". Hawkins, J. A. and H. the world. Holmback (eds.) Papers in Universal The author examines several models Grammar, special issue Lingua. oflan uage processing and de cides on the so-cal f ed Cohort Model, which claims, Benilde Grafia López among other things, that the recognition of a lexical item by the hearer occurs Universidad de Oviedo before the whole word is heard and the point of recognition depends on the extent to which the phonological form of the word is shared by other entries. For R. BOTHA, Chdlbnging Chomsky. example, on hearing the word trespass, a oxford: ~~~il ~l~~k~~ll. 1990. pool of competing candidates which share initial acoustic properties (i.e., a + 268 pages. "cohort") becomes activated. Thus, on reception of the first S, the cohort made The tide of this book will undoubtedly up of the words trespass, trestle and tress, arouse the interest of m-0 tyPes of readers: may be generated, and when the p is th0s.e who ~ractising -0r simply admireceived, the last two will drop out so that ring- Chomsban linguistics would be a unique item is isolated in he cohon- and eager to see what sort of challenge there recognition takes place. might be, wil-ling to confront it as if it ~h~ crucial point is that, as were directly aimed at them; and those complex words, prefKation entails more who being involved in different linguistic complexity than suffixation both for refiame-works would be delighted to presentation and recognition, in the case find in a book something they have at of prefixed words two different processes some time or 0th~ reflected upon. Ob- of generation of possible candidates and viously the first group will anticipate the subsequent selection are involved, one for evident superiority of the Master? theories the refix and one for the stem, whereas (to use Bothas own term) and the others in t R e case of suffixation a single process will be willing to applaud the critics' CO~- for the stem is enough, since, according ments as clearly more convincing. The to Hall, the sufix does not provoke the two groups will get their part, it must be generation of a new cohort because the said, because this book achieves an admirforms res~onding would be reduced to able balance in ~resentin~ the strongest the suffix itself, and a single member and the weakest aspects of Chomskfs cohort makes no sense. In short, the comviewpoint. Moreover, the style which Bobination prefix + stem involves a greater tha has chosen -presenting a confrontacost to the lexical processor and this is tion in the context of a game being played why it decides on sufixes rather than in different fieldsmakes his work both prefixes. amusing and appealing, without hiding