This book "supplies a vocabulary of English words and idiomatic phrases 'arranged … according to the ideas which they express'. The thesaurus, continually expanded and updated, has always remained in print, but this reissued first edition shows the impressive breadth of Roget's own knowledge and interests."
AUTHOR: Zeki Hamawand TITLE: The Semantics of English Negative Prefixes PUBLISHER: Equinox Publishing Ltd. YEAR: 2009
Galit W. Sassoon, ILLC, University of Amsterdam
This book presents research on a wide range of negative prefixes arguing that (i) each affix has multiple interpretations, prototypical and peripheral; (ii) the set of interpretations of an affix are partially overlapping with the interpretations of other affixes; (ii) due to a competition between affixes, a decision to use one rather than another always affects interpretation. Chapters 1-2 are devoted to criticism of previous theories and description of the framework, methodology (cognitive linguistics and usage based semantics), and main goals and ideas. Chapter 3 introduces a list of the interpretations for each affix. Chapter 4 defines DOMAINS - super-categories of interpretations; it describes the common denominator between prefixes belonging to a single domain, as well as different FACETS - subcategories of interpretations - along which these prefixes differ. Two prefixes and two domains are presented by way of illustration in the introductory chapters. Chapter 5 explains the notion of CONSTRUAL of a situation - the ways in which a situation is linguistically encoded. In this chapter, the author presents word pairs sharing a base but differing in prefix, such as 'non-social' vs. 'asocial', exemplifying his ideas about the prefixes and their categories.
All in all, the book describes eight general domains of interpretation for negative prefixes, and their different facets, as follows.
The domain of DEGRADATION contains affixes that can express decline in rank, importance, size, etc. (p14-15) where, typically, 'ab' expresses quality (as in 'abnormal'), 'sub' expresses degree (as in 'sub-human') and 'under' - rank (as in 'under-servant').
The domain of INADEQUACY contains affixes that can express partial possession of the thing described by the base (p36) where, typically, 'pseudo' expresses deception (as in 'pseudo-intellectual'), 'quasi' expresses resemblance (as in 'quasi-judicial role') and 'semi' - incompleteness (as in 'semi-literate').
The domain of DISTINCTION contains affixes that can express contrast and dissimilarity between entities or their properties. The prefixes in this domain differ along the scale of contrariness-contradictoriness. The most strongly associated with contrariness is 'in', followed by 'un', 'dis', 'a', and 'non', which is the most contradictory. 'Non' is the most neutral and least evaluative (followed by 'a', etc.), as in, e.g., 'non-professional' vs. 'unprofessional' (the former relates to people who are not trained, while the latter evaluates people as being under the standard expected in their profession), 'non-essential' vs. 'inessential' (roughly, not necessary vs. not important) and 'non-approval' vs. 'disapproval' (mere lack of approval vs. condemnation), 'atypical' vs. 'untypical' work (only the latter can mean 'bad'), 'non-moral' and 'amoral' vs. 'immoral', 'unmeasurable' (something for which no measuring means exist) vs. 'immeasurable' (too huge to be measurable), etc. (cf. p130-137 and notes 3-5 of ch.5).
The domain of OPPOSITION contains affixes that can express disagreement (the idea of 'in response to'), through reacting, preventing, defending, etc. (p104) where 'anti' is typically used to express opposition to attitudes and opinions (as in 'anti-colonialism'), harmful medical states (as in 'anti-histamine') and weapons (as in 'anti-missile'); 'counter' describes an event or action to deal with a situation (as in 'counter-attack') and 'contra' expresses extreme contrasts between things (as in 'contra-distinction'). Thus, we find the pair 'anti-revolution' (objection to the idea of a revolution) vs. 'counter-revolution' (revolution against the result of a prior revolution; p138-9).
The domain of TREATMENT contains affixes expressing wrong, false or improper treatment of people or things (p116-119), where 'mal' relates to intentional behavior (as in 'maladministration') and 'mis' unintentional or accidental behavior (as in 'misconstrued'); cf. 'maladjustment' vs. 'misadjustment' (inability to properly adapt vs. non-fit; the latter applies to, for instance, knobs) and 'maltreat' vs. 'mistreat' (roughly, treat cruelly or brutally vs. treat unkindly; p143).
The domain of PRIVATION contains affixes that can express lacking, not having, and preventing from having (p108). It is argued that typically 'dis' is used to express actions such as harming, reducing and punishing people (as in 'dishonor'), 'de' expresses damaging, neglecting or relocating of concrete things (as in 'debase') and 'un' relates to abstract things and situations (as in 'unhealthy'). While 'dis' conveys lacking, 'un' conveys 'bereft of', illustrated by the pair 'disbelief' (lack of belief) vs. 'unbelief' (lack of faith) and 'disburden' vs. 'unburden' (p139-140).
The domain of REMOVAL contains affixes that can express removal, releasing, ridding or taking away, where typically 'dis' relates to people (as in 'disarmed'), 'de' to things (as in 'degreased') and 'un' to objects ('unload'). In p140, we find 'decoupled' vs. 'uncoupled' and 'defrock' vs. 'unfrock' (remove from a privileged position vs. remove clothes).
The domain of REVERSAL contains affixes that can express reversing, turning around, or inverting from one state to the opposite (undoing), where typically 'dis' refers to people (as in 'discredit'), 'de' to things and places ('decipher') and 'un' to objects ('unlock'); cf. 'debar' (e.g., people from a bar) vs. 'unbar' (e.g., a door) and 'disvalue' (to think of something as having a low value) vs. 'devalue' (to lower the value of something; p141-2).
Finally, inter-domain construals are argued to be responsible for interpreting rival prefixes from different domains, as in 'non-human' vs. 'anti-human' (that relate to a category distinction vs. opposition with respect to humans), 'unfortunate' vs. 'misfortunate' (category distinction vs. treatment), 'displaced' vs. 'misplaced' (reversal vs. treatment), etc.
The main merit of this book lies in the attempt to study a wide range of negative prefixes comparatively. As argued in section 1.2, one does not find comprehensive comparisons between negative prefixes in, for example, dictionaries, reference grammars, and morphology. Furthermore, the author's basic assumption, namely that the interpretation of a prefix in a given context of use is affected by the interpretation of other prefixes, is interesting and plausible. In accordance, the book provides interesting insights regarding the interpretation of different prefixes. Nonetheless, some of the author's conclusions are not sufficiently convincing, due to problems with both the empirical procedure and the manner of presentation of the results.
First, the book contains long and detailed introductory descriptions of cognitive-linguistic and related theories, general critique of formal theories, descriptions of structure of the book and its parts, goals, procedures and summaries. These issues occupy a large part not only of chapters 1 and 2, but also of each one of the three main chapters (3-5). As a result, the core parts consisting of the actual study and its results cover less than a half of the whole book (mainly, pages 60-84 in ch.3, pages 99-120 in ch.4 and pages 130-148 in ch.5). This is a pity given the admirable, ambitious goal of providing a comprehensive study of a wide range of negative prefixes and the relations between them. It is a pity also because the 'introductory' sections are repetitive. For example, chapter 3 has two almost identical descriptions of its structure, in the first and second page of the chapter; so does chapter 4. Also, Prototype theory and Domain theory are presented at least 3 times. The sections criticizing previous formal theories cite theories whose nature and problems are well known, including the classical theory, rule-based morphology and referential semantics. Some of the cited theories are from 1930-1960 (p91, section 4.2.1). These reviews could have been dispensed with (by referring to any of the existing introductions to cognitive linguistics), leaving space for a review of contemporary literature about negation and antonymy, whether in psycholinguistics (such as Giora 2006 and references within), corpus linguistics (such as Tribushinina 2009 and references within) or formal semantics (such as Kennedy 2001 and references within).
Second, section 1.4.2 explains the merits of Usage-based semantics, according to which the linguistic system is shaped by actual data. Collocations and frequency of tokens and types are said to be the core factors (say, number of occurrences of 'un', number of occurrences of prefixes sharing a given interpretation with 'un', etc.) In fact the author looks at real examples taken from the BNC and their collocations. Yet, the book does not present any quantitative measure of frequency of occurrence or co-occurrence. The book provides lists of possible interpretations for each affix, illustrated by real examples. Yet, nothing reveals the extent to which these lists are exhaustive or the extent to which the generalizations are conclusive (how many of the examples in the BNC do they exhaust?). The author divides the prefixes into primary and secondary, and orders the different interpretations of any given prefix by prototypicality, but he remains silent as to the motivating evidence for each specific claim. The author disregards non-affixal negation and negative suffixes without justifying this choice (don't negative prefixes compete also with negative morphemes other than prefixes?). Finally, the author states for almost any example of a negative word what it means (e.g., 'mislay' means ''to lose something temporarily by forgetting where you have put it''; p117), based on dictionary definitions and personal intuitions, although the author himself states that these are non-reliable sources. It is certainly possible that the author has done very good and careful work, based on reliable, representative samples of examples, dictionaries and informants. However, in the absence of concrete details about these samples, the validity and generality of his claims remain in doubt.
Third, at times, it is not so easy to see the logical relations leading from data (examples, collocations) to conclusions (claims regarding different prefixes), in particular in chapters 4 and 5. For example, the author points out dissimilarities between the collocations characteristic of different words, e.g., 'decipher' collocates with 'messages' and 'wording', while 'deforest' with 'forest' and 'land' (p115), but what can we conclude here? The author could not point out a common denominator of the collocations of, e.g., 'de' words versus 'un' words in general, which is disappointing. The prefixes are also characterized in slightly different ways in different chapters and sections. Consider, for example, the domain of distinction. In p99-104 it is argued that, typically, 'in', 'ir' and 'im' are used to express properties of situations (as in 'inappropriate' and 'irregular'), 'un' is used to express properties of things ('untidy'), 'dis' - attitudes of people ('dishonest'), 'a' - opposing features of things ('atonal') and 'non' different plans of action ('non-compliance'). However, the picture drawn in chapter five is a bit different. For example, on p131, it is argued that 'a' applies to people and animates (as in 'he was asocial') and 'non' to areas of knowledge (as in 'non-social areas of linguistics'; cf. also to 'unsocial hours'). In p132, it is argued that 'dis' applies to people and things (as in 'disappearance') and 'non' to humans (as in 'non-appearance'). On p133, it is argued that 'un' applies to people's behavior (as in 'unprofessional') and 'non' to people (as in 'non-professional'). On p135 it is argued that 'a' applies to people and their attitudes while 'in' to their actions (as in 'amoral person' vs. 'immoral action'), and so on. To the best of my understanding, the different sections provide different, mutually inconsistent generalizations regarding the characteristic collocations of e.g. 'non'; likewise for other prefixes.
Also, despite the fact that minimal pairs such as 'non-social' vs. 'asocial' constitute the most straightforward evidence for differences between rival prefixes, the main discussion of such minimal pairs is delayed up to the last chapter and includes only one or two examples for each prefix pair. Again, in the absence of an exhaustive sample of examples, no generalization can be drawn. Additional invaluable examples are found in the notes, but the notes do not allow enough space to establish important arguments, and constant shifts from the chapter to the notes do not facilitate readability. As a result, the connections between the examples given in chapter 5 and the general claims presented in chapters 3-4 often remain loose.
Fourth, some of the terminology used in the book is not sufficiently explained. For example, what is the distinction between the terms 'thing' and 'object', as used in chapter 4? And more importantly, what is negation? What is a contrast? What is an opposition? What creates them? When is it the case that something is not? In order to provide a satisfying answer to these questions it is not sufficient to use many near-synonyms of negation (cf. p99). With these loose ends, certain facts remain unexplained; for example, it is argued that in 'unqualified' vs. 'disqualified', 'un' evokes the privation domain, while 'dis' - the reversal domain. But nothing explains precisely how and why this happens to be the case; according to the claims in ch. 3 there are many other possible scenarios.
At last, some explanations are redundant or suffer from flaws; to take one example, it is argued that the basic interpretation of 'semi', namely 'half' (as in 'semi-tone'), can change and extend into another, namely 'twice' (as in 'semi-weekly'). But no reference to extensions is necessary here. The typical interpretation ('occurs every half a week') already entails 'twice a week'. Stipulating an additional peripheral sense for 'semi' is unnecessary. To take another example, Horn (2002) argues that 'un' differs from 'non' with respect to the creation of contrary vs. contradictory antonym pairs (namely ones that allow or do not allow for a borderline area, as in e.g., 'tall/short' vs. 'open/closed', respectively). Expanding on this idea, the author claims that things can be contradictory or contrary to different degrees. However, he does not explicate precisely how these degrees are characterized and in several different places he argues that contradictoriness creates non-gradable properties, which is clearly wrong, e.g., 'open' implies 'not-closed', and yet things can be somewhat open, very open, too open, not open enough, etc.; likewise, examples such as 'typical/atypical' are used to illustrate a high degree of contradictoriness, although these adjectives are clearly gradable (see Kennedy and McNally 2005 for a lengthy discussion of these matters).
To conclude, this book can be useful and inspiring for readers working on the topic, but it cannot serve as a basis for firm conclusions regarding the semantics and usage of negative prefixes.
Giora, Rachel (2006). Is negation unique? On the processes and products of phrasal negation. Journal of Pragmatics, 38, 979-980. Horn, L. (2002) Uncovering the un-word: A study in lexical pragmatics. Sophia Linguistica 49: 1-64. Kennedy, Christopher. 2001. Polar opposition and the ontology of degrees. Linguistics and Philosophy 24(1): 33-70. Kennedy, Christopher and McNally, Louise. 2005. Scale structure and. the semantic typology of gradable predicates. Language 81: 345-381. Tribushinina, Elena, 2009, The linguistics of zero: A cognitive reference point or a phantom? Folia Linguistica 43/2: 417-461.
ABOUT THE REVIEWER
ABOUT THE REVIEWER:
Galit Weidman Sassoon is a postdoc researcher in the institute of logic,
language and computation (ILLC) at the University of Amsterdam (UVA). Her
main interests are natural language semantics and pragmatics, and their
interface with cognitive psychology. She currently works on problems
pertaining to nouns and adjectives, including vagueness, comparison and