Journal of Undergraduate Research
Volume 4, Issue 12 - September 2003
Empirical Support for a Minimalist Analysis of Zero Derivation
Stuart LaRosa
INTRODUCTION
Zero derivation is a phenomenon in which a word functions syntactically as a lexical category other than the one it is lexically assigned. It is common in English, but entirely absent in many other languages such as French or Japanese. Zero derivation in English is a productive process, and its uses run along a continuum from standard to innovative:
(1) The waiter will bottle the wine.
(2) Please Xerox these for the boss.
(3) The teacher’s gonna ruler your knuckles if you don’t cut it out!
Example (1) shows a zero-derived verb that has long stood as a standard use of zero derivation. Example (2) shows a zero-derived verb that has recently passed into standard usage. Example (3) shows a zero-derived verb that has not passed into standard usage. Note that (3) is entirely intelligible to a native speaker of English in spite of its innovative status.
In a 1999 University of Florida dissertation, Jo Tyler provided a comprehensive analysis of denominal verbs in English within the Minimalist paradigm. Her work, entitled The Syntax and Semantics of Zero Verbs: A Minimalist Approach, builds on the theoretical model created by Hale & Keyser (1992, 1993), which holds that the conflation structure of a denominal verb is essentially similar to that of the VP shell structure propsed by Larson (1988) for verbs with more than one internal argument. Figure 1 illustrates this similarity in a simplified structure.
Figure 1. Similarity between the Larsonian VP shell structure and Hale & Keyser's zero verb.
In a triadic verb such as put, a process of incorporation (head-to-head movement) moves V to v. The same process of incorporation works on a denominal verb to move N three times: first to P, then V, then ν.
The principal difference between Hale & Keyser’s analysis of denominal verbs and the one put forth by Tyler lies in their approach to the determination of thematic roles. Hale & Keyser attempt a configurationally defined approach to thematic relations, but Tyler points out that their model fails to adequately explain the difference between Theme and Goal, which occur in identical configurational contexts. To bring Hale & Keyser’s approach in line with Minimalism, Tyler proposes the feature [±central] as an inherent feature of the category Preposition, including abstract P as it occurs in a denominal verb’s conflation structure. Denominal verbs then fall into two categories based on the nature of their internal abstract P: structures including a [-central] P are Goal verbs (subsuming the classes Location verbs and Figurative Goal verbs) and those including a [+central] P are Theme verbs (subsuming the classes Locatum verbs and Instrument verbs).
Tyler theorizes that the [±central] feature of prepositions is a parameter of universal grammar. This feature, closely related to the notion of telicity, has been observed in numerous languages, including Warlpiri (Hale, 1986) and Romance (Mateu, 2001). William Croft conducted what is probably the most extensive research on this topic, identifying over 40 languages that mark arguments according to this feature. (Croft, 1991). Crucially, Melissa Bowerman has demonstrated that English-speaking children frequently confuse prepositions, but never use a [+central] preposition in place of a [-central] one or vice versa (Bowerman 1983). Thus, it is easily surmised that this particular feature does play a role in the structure of English, and critically so in the creation of denominal verbs.
By providing an entirely configurational approach to zero derivation, Tyler succeeding in bringing the phenomenon entirely in line with the principles set forth in Chomsky’s The Minimalist Program. This reasoning strikes me as sound, but since Tyler’s account is a fundamentally theoretical one, she provides primarily theoretical support for her proposal. What follows are the results of a study I conducted in the winter of 2002 that I hoped would demonstrate empirically the credibility of Tyler’s account.
METHODOLOGY
To test Tyler’s Minimalist account of English zero derivation, I devised a method to probe native speakers’ intuition about the primacy of words in noun/denominal-verb pairs. In the questionnaire’s directions, I explained briefly that some words in a language are older than others, and that in the case of words with multiple meanings, one meaning is often older than another (such as mouse “rodent” versus mouse “computer part”). I then asked the participants to think about a series of sentence pairs, each of which contained two uses of the same italicized word, and to choose which one they thought was the older of the two. Sentences (4 – 7) show examples of these pairs:
(4) The waiter will bottle the wine for us.
(5) The wine comes in a green bottle.
(6) The farmer built a pen for his pigs.
(7) The farmer told his son to pen the sheep.
In addition to the 20 noun/verb pairs I studied, I included in the research design a number of noun/noun pairs, verb/verb pairs, and noun/verb pairs that did not come into existence through simple zero derivation to help to prevent a learning effect among the participants. In each item, I used only bare, uninflected forms of both the noun and the verb. I made an effort to keep temporal references to a minimum; where they were necessary to keep a sentence from sounding stilted, I tried to keep both sentences in the same temporal frame of reference. All metonymic uses of the experimental words were eliminated to the best of my ability.
The questionnaire was administered to 50 18-year-old high school seniors in Tallahassee, Florida. Since Linguistics does not typically make up part of a high school curriculum, the participants were assumed to have little to no experience in linguistic theory. Additionally, these 18-year-olds had had no formal instruction about the history of the English language or its deeper machinations; any intuitions they had on the subject of word primacy, then, had to be based on some other area of knowledge, even if they felt they were making this judgment based on etymological history, as they were instructed to do. It is my belief that the only such knowledge they had available to them is lexical primacy. Since a denominal verb is syntactically derived from a lexical noun in Tyler’s model, the results of the questionnaire should reflect a preference for the nouns over the verbs based that primacy.
RESULTS AND CONCLUSIONS
Table 1 shows the results of the study.
| Table 1 Study Results |
||||||
Goal Verbs |
Theme Verbs |
|||||
|---|---|---|---|---|---|---|
| Exp. Word | N/V | %N | Exp. Word | N/V | %N | |
| Tree | 50/0 | 100 | Initial | 39/11 | 78 | |
| Bag | 49/1 | 98 | Poison | 39/11 | 78 | |
| Corner | 45/5 | 90 | Spear | 38/12 | 76 | |
| Bottle | 41/9 | 82 | Date | 36/14 | 72 | |
| Pen | 40/10 | 80 | Bomb | 34/16 | 68 | |
| Jail | 39/11 | 78 | Glue | 32/18 | 64 | |
| Knight | 35/15 | 70 | Oil | 30/20 | 60 | |
| Braid | 33/17 | 66 | Hammer | 30/20 | 60 | |
| Photograph | 32/18 | 64 | Bleach | 27/23 | 54 | |
| List | 29/21 | 58 | Bandage | 26/24 | 52 | |
| Total | 393/107 | 78.6 | Total | 331/169 | 66.2
Participants showed a preference—with varying degrees of strength—for the noun in all 20 of the experimental word pairs. Since they could not have been making this decision based on actual word history, this supports Hale & Keyser’s model of a verb syntactically derived from a lexical noun. Furthermore, participants made a marked distinction between Goal ([-central]) verbs and Theme ([+central]) verbs. On average, they chose the noun as “older” (primary) in 78.6% of the cases, versus 66.2% for Theme verbs. This clear divide supports Tyler’s hypothesis that a [±central] distinction plays a crucial role in the conflation structure of a denominal verb. DISCUSSION AND FURTHER ISSUESAs noted above, some experimental verbs, such as tree, bag and corner, were very readily identified as being “newer than” (i.e., derived from) their nominal counterparts, whereas other verbs, such as bandage, bleach and list elicited less clear-cut responses from the participants. Neither Hale & Keyser’s configurational approach nor Tyler’s revisions provide an explanation for why this should be so. An etymological review from the Oxford English Dictionary revealed no correlation between the words’ derivational history of accepted use and the participants' frequency of selection. In other words, it was not necessarily the case that denominal verbs that have been in standard use for a longer period of time were more commonly selected as being older than their nominal counterparts. A comparison with word frequency data might yield some insight into this issue, but that kind of data was not available to me. In my view, however, the disparities in frequency do not diminish the significance of the fact that the noun was chosen over the verb a majority of the time in all 20 cases. While it is not immediately clear which factors lead to Goal verbs being more readily understood to be derived than Theme verbs, this distinction is significant. More than anything, it strongly suggests that there is a true interpretable difference between [+central] and [-central] abstract P, an assumption on which Tyler’s model crucially rests. What remains unexplained, however, is why Goal verbs are more interpretable than Theme verbs and not vice versa. This is an interesting theoretical consideration that could provide insight into our understanding of this feature. In demonstrating zero derivation’s compatibility
with the Minimalist framework, Tyler has clarified not only many theoretical
questions surrounding zero derivation but also Minimalism itself. The
results presented above show that there is indeed some measurable difference
between Goal denominal verbs and Theme denominal verbs, a concept which
Tyler pioneered. A successful linguistic paradigm should be all-encompassing,
and Tyler’s work, supported here, has brought Minimalism one step
closer to being just that. REFERENCES
Bowerman, Melissa. 1983. Hidden meanings: The role of covert conceptual
structures in children’s development of language. In Rogers, D.
R. & Sloboda, J. A. (Eds.), The Acquisition of Symbolic Skills,
New York: Plenum, 445-470. Back to the Journal of Undergraduate Research
© University of Florida, Gainesville, FL 32611; (352) 846-2032. | |
