The company women and men keep: what collocations can reveal about culture

Studies in corpus linguistics (e.g. Partington 1998, Sinclair 1991, Sinclair 1996, Tognini Bonelli 2000) have demonstrated that words convey meanings especially as a result of their combinations with other words; thus, an analysis of their frequent co(n)texts of use helps characterize their denotational and connotional meanings, and assess their users’ attitudes to the referents and concepts they are associated with. In this work I compare collocations of common English words that encode the notion of gender: “woman – man”, “girl – boy”, and “female – male”. Data from the Cobuild on-line corpus show that such terms are not always used in the same co-texts. The top collocates of the ‘woman’ words fall within the domains of physical attractiveness (“pretty, beautiful, attractive”) or appearance (“older, hair, body, sex, sexual”), and family relations (“child, husband, married”). The semantic preferences of the ‘man’ words also include the domains of physical attributes — but to the exclusion of sexuality — (“younger, aged, dressed, haired, fat”) and family — but with less rich variety — (“marry”); in addition, these words include reference to non-physical attractiveness (“rich, nice”) and occasionally encode processes that represent men as agents (“standing, killing, arresting”). The collocational patterns of the ‘woman’’s and ‘man’’s words in selected syntactic environments also differ: - the former more frequently act as patients in passive constructions (i.e. next to past participles and “by”); when the latter appear in the same co-text, they are modified by adjectives identifying a subset of a class (“average, common”), or preceded by nouns referring to a part of the person identified (thus showing that this person is only partially affected by the process in question; “face, figure”); - the latter’s embedded descriptive prepositional phrases beginning with “of” are typically positively connoted by adjectives (“great, formidable, exceptional, immense, proven, rare”) and often refer to moral-cognitive virtues (“ability, courage, merit, principle, repute”); the former’s “of”-headed prepositional phrases may be negatively connoted (“dubious, fierce, terrifying”), often refer to beauty (“charm, beauty, profile”) or personality traits (“vitality, sweetness, temperament, determination”), and may introduce the possessive constructions of his own (which present the woman as a possessee or attribute of the man); - when part of prepositional phrases introduced by “for”, the ‘woman’ words are typically preceded by adjectives expressing judgments or deontic modality (“advisable, possible, suitable, necessary; need; acceptable, fit, unbecoming, unnatural”) and accompanied by words expressing negative notions (“problem, misery, guilt, concern, preoccupation; hard, not easy, absurd, incongruous”); the ‘man’’s words found in for-headed prepositional phrases, instead, are more often preceded by positive collocates (“admiration, affection, amusement, applause, not bad, far better, appropriate, convenient, easier, enough, fine, important, intense, right”) than negative ones (“hard, tragedy, treason, impossible, unnatural, trouble, shocking”) and followed by neutral or positive ones (“application, clothes, contemplating, cut, energy, perfection, position, prosperous, support, voice”). The findings indicate that the ‘woman’’s words are often enveloped in notions of passivity, negativity, and physicality, and that the ‘man’’s words, instead, more frequently appear to be vehicles for conveying ideas of activity, positivity, and cognitivity. This suggests that socio-culturally salient concepts encoded in potentially symmetric terms may develop different ideological significance when they recurrently keep company with different sets of words. References Partington, A. (1998), Patterns and meanings; Amsterdam: Benjamins Sinclair, J. (1991), Corpus, Concordance, Collocation; Oxford: OUP Sinclair, J. (1996), The search for units of meaning, Textus 9: 75-106 Tognini Bonelli, E. (2000), Corpus Linguistics at work; Birmingham: TWC