In generative grammar, a parasitic gap is a construction in which one "gap" appears to be dependent on another "gap", that is, the one gap can appear only by virtue of the appearance of the other gap, hence the former is said to be "parasitic" on the latter, e.g. Which explanation did you reject __1 without first really considering __2? While parasitic gaps are present in English and some related Germanic languages, e.g. Swedish (see Engdahl 1983), their appearance is much more restricted in other, closely related languages, e.g. German and the Romance languages. Japanese linguistic scholar Fumikazu Niinuma has attempted to differentiate between parasitic gaps and coordination in his research, as he believes the two are often confused. An aspect of parasitic gaps that makes them particularly mysterious is the fact they usually appear inside islands to extraction. Although the study of parasitic gaps began in the late 1970s, no consensus has yet been reached about the best analysis.
The following b-sentences illustrate typical parasitic gaps. The parasitic gaps are marked with a p-subscript:
- a. You reviewed that book without actually reading it. – No gap at all present
- b. What book did you review __ without actually reading __p? – Parasitic gap possible
- c. *You reviewed that book without actually reading __p. – Parasitic gap impossible without "real" gap
- a. They played that song repeatedly despite not liking it. – No gap at all present
- b. Which song did they play __ repeatedly despite not liking __p? – Parasitic gap possible
- c. *They played that song repeatedly despite not liking __p. – Parasitic gap impossible without "real" gap
- a. You bought that old bike in order to fix it up. – No gap at all present
- b. Which old bike did you buy __ in order to fix __p up? – Parasitic gap possible
- c. *You bought that old bike in order to fix __p up. – Parasitic gap impossible without "real" gap
The a-sentences are normal declarative sentences that contain no gaps at all. Each b-sentence, in contrast, contains two gaps, whereby the second gap is parasitic on the first. The c-sentences illustrate that if there is no "real" gap (that corresponds to the wh-expression in bold), the parasitic gap is not possible. One interesting thing about parasitic gaps like the ones here in the b-sentences is their motivation. Their appearance appears to be reliant on syntactic movement (e.g. wh-movement or topicalization). The fact, however, that there are two gaps in each b-sentence but only one fronted wh-expression is a source of the difficulty associated with the construction. How does it come to pass that one fronted wh-expression is capable of licensing two gaps? Another interesting fact about parasitic gaps is that they usually appear inside extraction islands (as they do in the examples just given), hence one might expect extraction from the site of parasitic gaps to be altogether impossible. The fact that the islands are ignored is a second source of challenge associated with the phenomenon.
Some historical notes
The phenomenon of parasitic gaps appears to have been discovered by John Robert Ross in the 1960s, but remained undiscussed until papers by Knut Tarald Taraldsen and Elisabet Engdahl explored the properties of the phenomenon in detail. The knowledge of parasitic gaps was central to the development of the GPSG framework (Generalized Phrase Structure Grammar) in the mid 1980s, this knowledge then being refined later in the HPSG framework (Head-driven Phrase Structure Grammar) of Carl Pollard and Ivan Sag. In the 90s, a debate centered around the best theoretical analysis of parasitic gaps (extraction vs. percolation), this debate culminating in a collection of essays edited by Peter Culicover and Paul Postal in 2001.
Some traits of parasitic gaps
The following subsections briefly explore some aspects of parasitic gaps that have been widely acknowledged in the literature on parasitism. The following areas are addressed:
- many parasitic gaps appear optionally;
- some parasitic gaps appear obligatorily;
- parasitic gaps appear in missing object constructions; and
- syntactic parallelism seems to promote the appearance of parasitic gaps.
Optional parasitic gaps
Many parasitic gaps appear optionally. They are in non-complementary distribution with a pronoun, meaning that the speaker has the choice whether to employ the gap or not, e.g.
- a. Which dish did you order __ after you tried __p? – Parasitic gap present
- b. Which dish did you order __ after you tried it? – Parasitic gap absent
- a. Which movie will they like __ as soon as they see __p? – Parasitic gap present
- b. Which movie will they like __ as soon as they see it? – Parasitic gap absent
The a-sentence contain typical parasitic gaps, whereas the b-sentence choose to use a pronoun instead of the gap. In other words, the parasitic gap in the a-sentences is occurring optionally. Optionality like this suggests an analysis of parasitism in terms of ellipsis, since optionality is the primary trait of known ellipsis mechanisms.
Obligatory parasitic gaps
While many parasitic gaps occur optionally as just illustrated, other parasitic gaps occur obligatorily. This can be the case when the parasitic gap precedes the "real" gap, e.g.
- a. The rumor about the girl annoyed her. – No gap at all present
- b. Which girl did [the rumor about __p] annoy __? – Parasitic gap precedes "real" gap
- c. ??Which girl did the rumor about her annoy __? – Real gap hardly possible unless the parasitic gap is present; weak crossover has occurred
- a. If you get to know him, you will like Bill. – No gap at all present
- b. Bill is the type of guy who [if you get to know __p], you will like __. – Parasitic gap precedes "real" gap
- c. ??Bill is the type of guy who if you get to know him, you will like __. – Real gap hardly possible unless the parasitic gap is present; weak crossover has occurred
These examples illustrate a couple of important facts about parasitic gaps. The b-sentences demonstrate that the parasitic gap can indeed precede the "real" gap, and the strong marginality of the c-sentences shows that in a sense, the real gap can also be dependent on the parasitic gap. Note that we know that the first gap (the leftmost gap) in the b-sentences is parasitic on the following gap because it, i.e. the leftmost gap, appears inside what is normally an extraction island (marked with square brackets). The aspect of parasitic gaps illustrated with these examples is addressed in terms of the weak crossover phenomenon (WCO). The WCO phenomenon occurs when a fronted expression is coreferential with an intermediate expression that appears between the fronted expression and the position of its gap. In the big picture, one can simply note that parasitic gaps behave variable depending upon whether they precede or follow the "real" gap. When they precede the "real" gap, their appearance is usually obligatory.
Missing object constructions
Much work on parasitism assumes that parasitic gaps are dependent on another gap, the "real" gap in the examples above. Hence the assumption is that parasitic gaps are reliant on those mechanisms that license normal extraction gaps, e.g. wh-movement and topicalization. This assumption is challenged, however, by so-called missing-object constructions (also known as tough-constructions), e.g.
- a. It is easy to appreciate her after getting to know her. – No gap present
- b. She is easy to appreciate __ after getting to know __p. – Parasitic gap present despite the lack of wh-fronting and topicalization
- a. It is hard to understand this essay without reading it several times. – No gap present
- b. This essay is hard to understand __ without reading __p several times. – Parasitic gap present despite the lack of wh-fronting and topicalization
- a. It will be tough to get the motor running without entirely rebuilding it. – No gap present
- b. The motor will be tough to get __ running without entirely rebuilding __p. – Parasitic gap present despite the lack of wh-fronting and topicalization
The a-sentences lack gaps entirely. The b-sentences contain parasitic gaps despite the fact that neither wh-movement nor topicalization has occurred. The b-sentences illustrate missing-object constructions, since the verbs appreciate, understand, and get are transitive and should hence take an object. This object is missing, as marked by the gap on the left. Whatever the analysis of parasitic gaps ends up being in the long run, it will have to accommodate the facts involving missing objects illustrated here. Movement (wh-movement, topicalization) may actually not be the key factor licensing parasitic gaps.
The role of parallelism
Examining the examples of optional parasitic gaps produced above so far, one sees that in each case, a certain parallelism is present. This parallelism is now illustrated using brackets:
- a. Which manuscript did you [resubmit __] after [revising __p]?
- b. Which foods does he [fantasize about __] without [ever eating __p]?
- c. Which report did you [file __] without [reading __p]?
- d. Which old bike did he [buy __] in order to [fix __p up]?
- e. Which girl did you [ask __ out] before [meeting __p in person]?
In each of these examples, the square brackets mark what appear to be parallel structures, as associated with the coordinate structures of coordination. The brackets mark verb phrases (VPs), whereby the subordinator appearing between the brackets is functioning like a coordinator (i.e. and, or, or but). This parallelism may be a significant factor that is aiding the appearance of the parasitic gaps. When this parallelism is absent, there is a significant drop in acceptability of the parasitic gap:
- a. ?Who [ __ secretly supports John] without [John secretly supporting __p back]?
- b. ?Which girl [ __ likes Billy] without [Billy liking __p back]?
- c. ?Which spy [ __ escaped] without [anyone first identifying __p]?
- d. ??Which explanation [ __ had to be repeated] for [us to finally get __p]?
- e. ??Which report [ __ was filed] without [any of us first reading ___p]?
These instances of parasitic gaps are all marginal to varying degrees. The marginality is probably due to the lack of syntactic parallelism indicated by the brackets, the gaps no longer appearing on the same side of the brackets. In any case, there is a noticeable drop in acceptability when the parallelism in the examples further above is removed. What exactly explains this drop in acceptability is not entirely clear, although it may have to do with ease of processing. Parallel structures are easier for humans to process, and hence parasitic gaps are reliant on a low processing load.
The theoretical analysis of parasitic gaps is not a settled matter by any means, since accounts of the phenomenon vary drastically. In very broad terms, there are two lines of analysis that one can pursue. The first is to assume that parasitic gaps are extraction gaps (the extraction analysis); parasitic gaps arise by way of the same basic mechanism that licenses "normal" extraction gaps. This sort of approach must augment the analysis of extraction gaps in some way in order to accommodate parasitic gaps under the same theoretical umbrella. The alternative approach rejects the analysis that takes parasitic gaps to be extraction gaps. One assumes instead that parasitic gaps actually contain a covert element, this element having the status of definite proform. (the proform analysis). Some analyses mix and match these two basic lines of analysis, although in general, both are well represented in the literature on parasitism and most accounts can be placed in the one or the other camp.
Extraction analyses have the advantage that they immediately accommodate the simple observation that most parasitic gaps appear to be dependent on the occurrence of wh-movement or topicalization. Extraction analyses are challenged, however, by missing-object constructions, as noted above. Proform analyses have the advantage that they immediately accommodate the simple observation that many parasitic gaps occur optionally; the covert proform has the option to be overt. The proform analyses are challenged, however, by the fact that most parasitic gaps occur in the immediate environment of wh-movement or topicalization, since they do not provide a clear basis for explaining this correlation.
- Parasitic gaps have been studied most using English and Swedish data. See Engdahl's seminal article (1983) in this regard.
- Fumikazu, Niinuma (2010). "Across-the-Board and Parasitic Gap Constructions in Romanian". Linguistic Inquiry. 41 (1): 161–7. doi:10.1162/ling.2010.41.1.161.
- See the introduction in Culicover's and Postal's collection of papers (2001) for an overview of the varying theoretical accounts of parasitic gaps.
- See Ross' seminal dissertation (1967/86) in this regard.
- See especially Engdahl's 1983 article in this regard – more than any other paper, this one got the ball rolling.
- In addition to the essays, Culicover's and Postal's book also contains an extensive overview of earlier accounts of the phenomenon.
- Concerning the importance of weak crossover to the theory of parasitic gaps, see for instance Engdahl (1983:17ff.), Culicover (2001:32f.), and Levine and Hukari 2001:194).
- The ability of missing-object constructions to license parasitic gaps is widely acknowledged, e.g. Engdahl (1983:12f.), Postal (2001:257), Culicover (2001:34).
- The role played by syntactic parallelism in determining the distribution of parasitic gaps has been explored by many, e.g. Williams (1990), Munn (2001), Culicover (2013:153ff.).
- For two examples of the extraction analysis of parasitic gaps, see Contreras (1984) and Chomsky (1986).
- For examples of the proform analysis, see Cinque 1990, Fiengo and May 1994, Postal 1994).
- Chomsky, N. 1986. Barriers. Cambridge, MA: The MIT Press.
- Cinque, G. 1990. Types of Ā-dependencies. Cambridge, MA: MIT Press.
- Contreras, H. 1984. A note on parasitic gaps. Linguistic Inquiry 15, 698–701.
- Culicover, P. 2001. Parasitic gaps: A history. In Parasitic Gaps, ed. by P. Culicover and P. Postal, 3–68. Cambridge, MA: The MIT Press.
- Culicover, P. 2013. Grammar and complexity: language at the intersection of competence and performance. Oxford: Oxford University Press.
- Culicover, P. and P. Postal (eds.) 2001. Parasitic gaps. The MIT Press.
- Engdahl, E. 1983. Parasitic gaps. Linguistics and Philosophy 6, 5–34.
- Fiengo, R. and R. May 1994. Indices and identity. Cambridge, MA: MIT Press.
- Levine, R, T. Hukari, and M. Calcagno 2001. Parasitic gaps in English: Some overlooked cases and their theoretical implications. In Parasitic Gaps, ed. by P. Culicover and P. Postal, 181–222. Cambridge, MA: The MIT Press.
- Munn, A. 2001. Explaining parasitic gap restrictions. In Parasitic Gaps, ed. by P. Culicover and P. Postal, 369–402. Cambridge, MA: The MIT Press.
- Postal, P. 1994. Parasitic and pseudo-parasitic gaps. Linquistic Inquiry 25, 63–117 [Reprinted in 2001 in Parasitic Gaps, ed. by P. Culicover and P. Postal, 253–313. Cambridge, MA: The MIT Press].
- Postal, P. 2001. Further Lacunae in the English parasitic gap paradigm. In Parasitic Gaps, ed. by P. Culicover and P. Postal, 223–253. Cambridge, MA: The MIT Press.
- Ross, J. 1967. Constraints on variables in syntax. Ph.D. Dissertation, MIT.
- Ross, J. 1986. Infinite syntax! Norwood, NJ: ABLEX [Reprinted dissertation from 1967].
- Williams, E. 1990. The ATB theory of parasitic gaps. The Linguistic Review 6, 265–279.