Portal de Periódicos da CAPES

Revisiting the passive of infinitival perception verb complements

2004; Routledge; Volume: 76; Issue: 1 Linguagem: Inglês

10.1080/00393270410027048

1651-2308

Dirk Noël,

Natural Language Processing Techniques

Click to increase image sizeClick to decrease image size Notes An example of a sentence that cannot be passivised is My mother had two horses: *Two horses were had by my mother. The research reported on in this paper was made possible by the Research Fund of the University of Ghent (Bijzonder Universitair Onderzoeksfonds contract nos. 12052095 and 12050399). But see Kirsner and Thompson Citation1976: 209–211; Huddleston Citation2002: 1237. (1b) is generally assumed to be impossible, but Bolinger (Citation1974: 86) and Duffley (Citation1992: 45) do not exclude it. Both agree, however, that it is extremely rare. This is an example plucked from the Internet: (i) The projected image changes to a collage of photos, taken from various spots outside a rickety old house, with various figures passing and going, inside and out. “This is our target. We have judged our intelligence to be correct, due to the large number of people coming and not being seen leave for a good amount of time. Obviously, there's got to be some type of substructure. We've heard about their ‘Underground underground base’ for a while, now. But we think we've found it.” (http://www.trailing‐edge.com/~cchang/arm/LOG‐UGRAID.HTML) All examples followed by a source code are taken from the second version (the so‐called “World Edition”, released in December 2000) of the 100‐million‐word British National Corpus (BNC). For a description of the corpus, see Aston and Burnard (Citation1998). In the source code, the three positions before the space identify the excerpt it was extracted from (which can be looked up in Burnard Citation1995); the number following the space is the line number of the example within the excerpt. In example (6) the to‐infinitival perception verb complement refers not to state but to an activity (i.e. a dynamic situation extended in time and under the control of an agent; cf. Lyons Citation1977: 483). This is only possible with progressive to‐infinitives, however. It is not accidental that the infinitive in such cases is to be, the most frequently occurring infinitive in accusative and infinitives. A semantic feature shared by activities and states is their continuity. I do not agree with Kirsner and Thompson's acceptability judgement of (17a), which can be compared with (i). (i) The Coke was cold and Harry drank it gratefully, even though he could feel that it brought him out in an instant sweat. (ECU 2506) http://www.catholicexchange.com/church_today/message.asp?message_id=2975&sec_id=3 Barwise (Citation1981: 373) distinguishes between “epistemically neutral” and “epistemically positive” perceptual reports, illustrating the former with (i) and (iii) and the latter with (ii) and (iv) (i) Dick saw Rosemary remove the crucial part of the Watergate tape. (ii) Dick saw that Rosemary removed the crucial part of the Watergate tape. (iii) Ralph saw a spy hiding a letter under a rock. (iv) Ralph saw that a spy was hiding a letter under a rock. For Barwise, (ii) and (iv) entail knowledge, while (i) and (iii) do not. Though there is a tendency, says Barwise (Citation1981: 374), for the hearer of (i) and (iii) to infer (ii) and (iv), respectively, such an inference is based not on the semantic content of (i) and (iii), but on a Gricean conversational implicature, which like all implicatures can be cancelled, as in (v). (v) Ralph saw a spy hiding a letter under a rock, but thought she was tying her shoe. One could argue, however, that if there is a tendency for the hearer to equate (i) with (ii) and (iii) with (iv), this equation has a higher reality value for the language user than the necessary outcome of a philosophical exercise. Barwise (Citation1981: 376) admits this himself when he writes: “In normal usage, NI [naked infinitive] perception statements are taken as veridical.” The distinction between different‐order entities is borrowed from Lyons (Citation1977: 442–445). A (set of) individual (s) is a first‐order entity, which can be located in space and evaluated in terms of its existence. A state of affairs is a second‐order entity, which can be located in space and time and evaluated in terms of its reality. A propositional content is a third‐order entity, which cannot be located in space or time and can be evaluated in terms of its truth. All possible forms of SEE were looked for, either immediately followed by the particle to or separated from it by up to four words. The frequency count in Noël (Citation2001) was based on a BNC query for occurrences of the past participle form of the verb either immediately followed by to be or separated from it by a by‐phrase with a one‐, two‐, three‐ or four‐word NP. BE seen to was not part of that study. Only the combinations of BE seen to with be produced by the query mentioned in the previous note were taken into account to compare frequencies. Both queries are slightly different in that the one used in this study does not mention by‐phrases, but since cases in which the patterns are interrupted by intervening material are not very numerous, this will not have affected the position of be seen to in the top 10. The index value in the last column was arrived at by dividing the absolute number of passive occurrences by the absolute number of active occurrences and should therefore be read as follows: “there are more than 35 times as many passive matrices containing the verb report than there are active matrices with this verb.” The unusually high percentage of lexical infinitives after presume should be treated with caution: 1 in 3 means exactly that, i.e. there were three cases of active presume followed by a to‐infinitive and one of them was a lexical verb. For the motivation of the order in which these patterns are listed, see Noël (Citation2001: 287–288). BE supposed to is not listed because unlike the others it has first and foremost grammaticalised into a deontic modal rather than an evidential. http://www.bermuda‐triangle.org/Theories/Electromagnetism/Behind_the_Controls/Lost_Aircraft/Navajo_­N3527E/navajo_n3527e.html http://www.nachos.net/vbull/content/topic/3727‐1.html http://bspace.freeshell.org/Stories/FishPart2.txt http://www.wildlifewebsite.com/beagle/beagle256.shtml http://www.talana.com.au/talana_rayne.htm http://www.imps4ever.info/racing/rallycross.html