Download English Conditional Inversion: A Construction-Based Approach and more Exercises English in PDF only on Docsity!
English Conditional Inversion: A Construction-Based
Approach
June 13, 2011
Abstract Conditional sentences also can be formed by inversion of subject and auxiliary, but it hap- pens only in a limited environment. This paper addresses grammatical constraints in condi- tional inversion and how they behave differently from the regular conditional clauses based on corpus investigations. Our corpus search reveals many different types of conditional inversion constructions, indicating the difficulties of deriving inverted conditionals from movement op- erations. In this paper, we provide a construction-based approach to the inverted conditional construction. The paper shows that the most optimal way of describing the general as well as idiosyncratic properties of the inverted conditional constructions is an account in the spirit of construction grammar in which a grammar is a repertory of constructions forming a network connected by links of inheritance.
Keywords: conditional inversion, construction-based, inheritance, counterfactual, corpus-based
1 Introduction
English allows many different types of subject-auxiliary inversion (SAI) constructions, as exempli- fied by the following attested examples (Green 1980, Quirk et al. 1985, Biber et al. 1999):
(1) a. Can you believe what he said? (Yes-no question)
b. Never had he more enjoyed spending money. (Negative inversion)
c. He has read more articles than have his classmates. (Comparative inversion) ∗An earlier version of this paper was presented at the 56th Annual Conference of International Linguistic Association, held at Rutgers University, April 15–April 17, 2011. I thank the co-presenter Sulki Park and the audiences of the conference for questions and suggestions. Our thanks also go to the anonymous reviewers of this journal for helpful comments. All errors are of course mine.
d. May he live to be one hundred and twenty with the eyesight of a falcon! (Wishing)
e. Don’t you even touch that! (Emphatic negative inversion)
f. Had I had enough money I would have bought a new car. (Conditional Inversion)
All these inversion constructions involve the repositioning of the auxiliary in front of the sub- ject, which is traditionally termed as I-to-C movement or SAI constructions (cf. Geis 1985, Haege- man 2009, 2010). Of these varieties of inversion examples, this paper concerns the conditional inversion construction exemplified in (1f). This conditional inversion construction has at least three different sub-types with respect to the auxiliary type in the inverted clause:^1
(2) a. Had she told him she was a princess, Peter could not have been more impressed. (COCA: 2005 FIC)
b. Were I eighty years younger, I might be honestly tempted by you. (COCA 2009 FIC)
c. Should gas rise above $4.85, the producers would have to repay the Treasury. (COCA: 2002 NEWS)
In terms of semantics, (2a) and (2b) are counterfactual in the sense that the inverted clause denotes a proposition that the speaker believes to be false whereas (2c) is just contingent in the sense that the occurrence of the main clause situation depends on the inverted clause (cf. Iatridou 2000, Iatridou and Embick 1994, Huddleston and Pullum 2002). In this paper, we try to answer the following research questions related to these three types of conditional inversion:
- What are the grammatical properties of the conditional inversion construction?
- Are there any grammatical constraints in using the conditional inversion?
- What are the relationships between the canonical if-conditional and conditional inversion constructions?
- Are there any relations with other inversion constructions?
In doing so, we will first look into basic properties of the conditional inversion construction, and see if our corpus data reflect the properties that the literature has discussed so far. We then sketch a construction-based, lexicalist view to provide an answer to the question regarding the relationship of the conditional inversion construction with other SAI constructions.
(^1) The corpora we have used in this study include 410 million words COCA (Corpus of Contemporary American English), 400 million words COHA (Corpus of Historical American English), and 100 million words BNC (British National Corpus). All these are available online.
(5) a. If that had happened, what would the economic look like today? (COCA: 2011 NEWS)
b. If I had not started this program, I would probably be depressed like my friends. (COCA: 2010 ACAD)
In such examples, the if-clause describes an unreal event, but the main clause tells us a potential outcome at a not past but present point. Note that unlike this type of counterfactual inversion, there is also another conditional where a subjunctive appears in the if-clause:
(6) a. If you were my girl, I would give you the best one in my life. (COCA: 2006 SPOK)
b. If no one were around, he would even pass the red ones. (COCA: 1993 FIC)
Just like examples in (5), these examples also mark a situation as not actually being the case at the time of speaking.^3 The if-clause can also include an auxiliary verb like should:
(7) a. If he should leave her, she would no doubt be telling someone in her whining voice. (COCA: 1995 FIC)
b. If anyone should be ready, it should be them. (COCA: 2003 NEWS)
These conditionals in (7) are not counterfactual: the auxiliary should places a strong contingent relation between the if-clause and the main clause.
2.2 Formal Properties of the Conditional Inversion
As noted, conditional sentences also can be formed by the inversion of subject and auxiliary, but it happens only in a limited environment. That is, conditional inversion is possible only with had, were, and should, not licensing inversion examples like the following (Newmeyer 1998, Fillmore 1999):
(8) a. If he has seen the film, I wouldn’t be shocked.
b. *Has he seen the film, I wouldn’t be shocked.
(9) a. If she was standing here with me, he would tell you the truth. (^3) The subjunctive verb were is nowadays often replaced by was:
(i) If I was a girl, I wouldn’t want to be a football player because everyone would be yelling at me or would look at me in a strange way. (COCA: 1992 NEWS)
b. *Was she standing here with me, he would tell you the truth.
(10) a. If she would plead guilty and agree to testify against her husband, the state would recommend it.
b. *Would she plead guilty and agree to testify against her husband, the state would recommend it.
The inverted conditional clause functions as a modifier as observed from its distributional pos- sibilities: it can appear in sentence initial, medial, or final as attested from the following corpus examples:
(11) a. [Had they done something], this animal would not have taken my family. (COCA 2009 SPOK)
b. We, the judges, [had we been the legislators], would have found such an intention sensible or morally or politically desirable. (BNC: GWN 177)
c. I should probably never have known either of the team any better [had not misfortune fallen upon them]. (COHA 1932 MAG)
These distributional possibilities support the modifier status of the inverted conditional clause (Bhatt and Pancheva 2005). The conditional inversion can also appear in embedded clauses:
(12) a. I felt that [had I done that], I would have been satisfying something that was there. (BNC K6M 349)
b. Her tone suggested that [had she been close enough] she would have elbowed him in the ribs. (COCA: 2009 FIC Bk:SheShootsConquer)
c. There never has been a year since my first nomination in which I could not have made a million [had I taken the side of privilege and favoritism]. (TIME 1925)
As the data illustrate, the conditional inversion can appear within the CP or even within the relative clause, indicating that the conditional inversion is not a main clause phenomenon. An intriguing constraint also exists in the inverted auxiliary. It cannot be Neg-contracted, unlike the canonical SAI construction (cf. Iatridou and Embick 1994, Bhatt and Pancheva 2005):
(13) a. Hadn’t we all suffered?
b. Weren’t they in my name?
(14) a. *Weren’t weather always perfect, they would be unable to give or express life.
(17) a. They would have been eligible for Tony consideration [only if] they had transferred to a Broadway theater. (COCA: 1994 NEWS)
b. Africans would only be given the right to reside permanently in a town or city [only if] they had been born there. (COCA: 1998 ACAD)
As such, the only if tells us an exceptional case required for the event of the main clause to happen. Slightly differently, when the exceptional condition expresses a possibility that has no effect on the main clause, we can use even if:
(18) a. [Even if] he had shown up there, it would not have mattered. (COCA: 1996 NEWS)
b. [Even if] I had asked, that wouldn’t have meant anything. (COCA: 2002 FIC)
Inverted conditional clauses also behave differently from the regular conditional clauses. For example, the non-inverted conditional antecedents may be modified by focus adverbs such as only and even, but this is not possible with inverted conditional antecedents according to Iatridou and Embick (1994):
(19) a. [Only if] Peter had come would Susan have left.
b. *[Only] had I thought that would I have called him.
(20) a. [Even if] she had been allergic to dill, he would have served the stuffed grape leaves.
b. ?(?) [Even] had Joe served truffles Kathy would not have been happy.
However, our corpus search gives us many examples where the degree focus expression even appears in conditional inversion:
(21) a. But the new rule, even had it been in effect, wouldn’t have helped Rosser. (COCA 2002 NEWS )
b. Even had I mustered the nerve to speak of it, nothing would have happened. (COCA 2005 FIC)
c. Even had she known his birthday, she wouldn’t have scoured Albuquerque thrifts. (COCA 2006 FIC)
The reason for the unacceptable conditional inversion with the focusing marker only can be attributed to the fact that the inverted clause conveys old information. As noted by Iatridou and Embick (1994), there are many environments where only cannot be in the focus position:
(22) a. He (only) admits (only) that he stole the tapes.
b. He heard the rumor (*only) that Bill had stolen the tapes.
A variety of the conditional inversion examples we have identified from the corpus show us that the subject is either a pronoun or a definite NP:
(23) a. Had [we] not packed snowshoes, we could not have followed its trail. (COCA: 2008 MAG)
b. Had [my father] succeeded in becoming a Catholic priest, I would still have been born. (COCA 2010 FIC)
(24) a. Were [I] more rational, I would be more certain about it.
b. Were [Karl] alive today, he’d be frantically updating it.
Unlike these counterfactuals, the contingent inversion allows its subject to be an indefinite:
(25) a. Should [people] be concerned, what should they do? (COCA 1999 SPOK)
b. Should [gas] not be available to supply these plants, it could trigger a serious economic crisis. (COCA 2000 ACAD).
Even though most of the contingent inversion examples found in our corpus data have a pronoun as the subject, we also find examples like (25) where the subject is not definite. The fact that the counterfactual inversion represents old information can also be supported from it-cleft examples. The regular if-conditional clause can be clefted, but the inverted clause cannot (cf. Iatridou and Embick 1994):
(26) a. It is [if John had come] that Mary would have left.
b. *It is [had John come] that Mary would have left.
The regular if-conditional clause can also be an answer to a wh-question, but not the inverted one:
(27) A: When/under what circumstances would Mary have come?
a. B: If she had been offered many artichokes.
b. B: # Had she been offered many artichokes.
These facts further support the constraint on the information structure for the inverted conditionals.
3 A Construction-Based Perspective
3.1 Form and Functional Constraints
The if-conditionals and conditional inversion data we have seen so far indicate the difficulties of deriving inverted conditionals from movement operations. In this paper, we provide a construction- based approach to the inverted conditional construction. As we have seen earlier in (1), English employs many different SAI (Subject-Aux Inversion) constructions as given in (33) (cf. Kim and Sells 2008):
(33) (^) SAI
eeeeeeeeeeeeeeeeeeeeeeeee
~~ ~~ ~~ ~~ ~~ ~~
~~ (^33) (^33) (^33) (^33) (^33) (^33) (^33) (^33) (^33) (^33) (^33) 3 OOO OOOO OOO OOO OOOO OOO OOOO OOO [[[[[[[[[[[ [[[[[[[[[[[ [[[[[[[[[[[ [[[[ Y/N Q Conditional Inv Neg Inv Emphatic Neg Inv Comparative Inv Wishing These constructions have one shared formal property such that the auxiliary is inverted in the sentence initial position as have been accepted by traditional grammar and others. However, the remaining question is, as Fillmore (1999) and Goldberg (2006) have suggested, if there are other shared functional properties. The set of SAI constructions, linked or derived from the prototypi- cal sentences, are different from prototypical sentences in many respects. For example, the SAI constructions cannot stand alone, are dependent upon the main clause, and denote non-positive assertions. Considering the typical conditional inversion in (34), we can easily identify these non- prototypical properties: (34) a. Had they known the cases, they might have acted differently. b. Were you in our shoes, you would show the same prudence. c. Should you leave the company, these funds are yours. The inverted conditional clause in these counterfactuals (34a) and (34b) as well as in the contingent conditional in (34c) are all non-declarative speech acts and do not assert or presuppose the truth of the proposition. That is, the conditionals here do not assert or presuppose that they knew the cases, you are in our shoes, or you leave the company. The antecedent clauses here cannot stand alone, depending upon the main clause. SAI constructions display the three common attributes: non- positive, non-asserted, and dependent. These common attributes on the SAI can be represented as constructional constraints in Figure 1: Sem/Prag: nonpositive & nonassertive Syntax: [ HEAD | INV+ ] 1 XP YP[SUBJ〈 1 XP〉] Figure 1: SAI in English The constructional constraints in Figure 1 specify that in terms of semantics and pragmatics, SAI constructions have non-positive and non-assertive illocutionary force. In terms of syntax, SAI constructions are inverted and have three sisters consisting of an inverted auxiliary, subject XP, and the predicate, as represented in the following tree structure: (35) (^) S mmmmmmmmmmmm TTTT TTTTT TTTTT TTTTT TTTT ### S ### MOD〈 5 〉 ### INV+ mmmmmmmmmmmm JJJ JJJ JJJ JJJ J ### 5 S ttt ttt ttt ttt ttt ttt JJJ JJJ JJJ JJJ JJJ JJJ ### V MOD〈^ ### INV+ (^) NP VP ~~
~~
~ @@@ @@@ @@ @@ (^) @ they might have acted differently. Had they known the cases, Even if the conditional inversion inherits the common attributes of the SAI constructions con- straints in Figure 1, as we have seen earlier, the conditional inversion has its own constructional constraints. That is, as we have seen, the auxiliary form values in conditional inversion are re- stricted to were, had and should and the inverted clause modifies the main clause. In addition, the auxiliary cannot accompany the contracted negation n’t. These morphosyntactic as well as condi- tional semantics are non-prototypical properties of the SAI, but are the properties of conditionals by themselves as stated in Figure 2.^6 The conditional inversion constructions, inheriting the properties from the SAI, will thus also observe the constraints in Figure 2. These constraints basically will allow us not to generate exam- ples like the following: (^6) As illustrated in Kim and Sells (2008), the VFORM value in English is canonically divided into two subtypes finite and nonfinite. The former type has three subtypes pres (present), past, and plain whereas the latter type has as its subtypes base, prp (present-participle), psp (past-participle), and inf (infinitive). In addition to these classifications, we we introduce had, should, were as subtype VFORM values of the past. a. 〈had〉 HEAD ### AUX + ### NEG – VFORM had ### MOD〈 (S) 〉 b. 〈has〉 HEAD [ AUX + VFORM has ] ### MOD〈 〉 The differences between had and has thus come from the feature value of NEG and MOD. While has has neither NEG nor MOD value so that they can appear in the main clause and with the contracted negation, the inverted had cannot be negated and need to modify a main clause. In British English, it is often assumed that the main verb have has also an AUX feature: (41) a. He had a lot of money. b. Had he a lot of money? We then expect a similar possibility in conditional inversion. In fact, this expectation is born out from our corpus data: (42) a. Had he a choice, he would never have chosen the life struggle of morbid obesity. (COCA 1992 NEWS) b. Had he his druthers, he would have liked to forge a new era of human-animal relations in Dr. Harmony’s fat butt. (COCA 2002 FIC) In addition, note that the modified main clause can be not only a declarative, but also an inter- rogative or imperative: (43) a. Had they tried that route, how far back would he have gone? (BNC: CGL 1754) b. Were he to do it, would it have an impact on the people voting in the House? (COCA: 1993 SPOK) c. Should she open it, what will she find? (COCA: 2008 FIC) (44) a. Should it not clear in a few days, make an appointment with your gynecologist. (COCA: 1993 MAG) b. Should you get an infection, check in with your gyno so she can diagnose and treat it ASAP. (COCA: 2008 MAG) These constructional constraints specified here are thus sui generis. That is, we cannot predict these properties from other general principles. Note that we need to subclassify the inversion con- ditionals into two subconstructions: counterfactual and contingent. One clear difference we have observed is that the inverted clause in the counterfactual conditionals denotes given information. This explains why the counterfactual conditionals do not occur in the it-clefts and other places designated as focus positions, whose data we repeat here again: (45) a. *It is [had John come] that Mary would have left. b. It is [if John had come] that Mary would have left. To reflect this information packaging constraint on the conditional inversion, we are led to assign this as a low-level constraint to counterfactual inversion constructions in Figure 3: Inherit SAI & Conditional Inversion Sem/Prag: given information Syntax: Figure 3: Counterfactual Inversion As indicated here by the empty values in the three sisters in Syntax, the counterfactual inversion construction inherits its syntactic information from SAI and conditional inversion constructions but has its own semantic and pragmatic constraints. ## 3.2 More on Construals As pointed out by Haegeman (2010), it has been proposed in literature (Larson 1990, Declerck 1997, among others), temporal clauses can have both high and low construals: (46) a. I saw Mary at Seoul when [she claimed [that she would leave]]. b. High construal: at the time that she made that claim c. Low construal: at the time of her presumed departure d. Had I not delved into my own emotions, I doubt the book would have resonated so deeply with readers. These examples are expected given that the conditional inversion basically represents ‘given’ infor- mation that can function as a topic element. As is well known, topic-like adverbials can be linked to the embedded clause (cf. Hukari and Levine 1995): (51) a. [On Monday], I think that [Kim went home very late]. b. [Yesterday], it seems that [Kim arrived home very early]. c. [How often] do you think that [Fred was late this week]? Whether we posit a movement operation for such or not, we can provide a parallel analysis for conditional inversion without positing an ad hoc operator like ‘world’ operator as assumed in Bhatt and Pancheva (2005). 4 Conclusion As we have seen, the device of constructional inheritance enables us to represent the English gram- mar as a ‘repertory’ of constructions while acknowledging significant grammatical generalizations in SAI as well as conditional inversion. We have first identified the proto-typical properties of the SAI constructions, different from canonical non-inverted sentences. We have also teased out the properties of conditional inversion constructions whose properties are partly inherited from their counterparts if-conditionals. The counterfactual and contingent inversion thus share some properties with its supertype SAI as well as if-conditionals, but at the same time have their own constructional constraints which are not pre- dicted from these related, high-level constructions. We have seen that this inheritance mechanism within the construction grammar perspective can provide an optimal way of describing the general as well as idiosyncratic properties of the inverted conditional constructions. That is, the inheritance mechanism allows us to capture the family resemblance among SAI constructions while mid- as well as lower-level generalizations are captured by sub-regularities on their sub-constructions. This way of explaining English grammar is an attempt to show the commitment in principle to account for the entirety of the language in question. References Bhatt, Rajesh, and Roumyana Pancheva. 2005. Conditionals. In M. Everaert, and H. van Riemsdijk (eds.), The Blackwell Companion to Syntax 4: 639-687. Blackwell. Biber, Douglas, and Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. MA: MIT Press. Declerck, Renaat. 1997. When-clauses and temporal structure. London: Routledge. Fillmore, Charles. 1999. Inversion and constructional inheritance. In Gert Webelhuth, Jean-Pierre Koenig and Andreas Kathol (eds.), Lexical and Constructional Aspects of Linguistic Expla- nation: Studies in Constraint-Based Lexicalism, pp. 113-128, Stanford: CSLI Publications. Geis, Michael L. 1985. The Syntax of Conditional Sentences. In G. Michael (ed.), Studies in Generalized Phrase Structure Grammar 130-159. Columbus, OH: Department of Linguistics, OSU. Goldberg, Adele E. 2006. Grammatical Categorization: Subject-Auxiliary Inversion. Construc- tions at Work: The Nature of Generalization in Language. NY: Oxford University Press. Green, Georgia M. 1980. Some Wherefores of English Inversions. Language 56: 82-601. Haegeman, Liliane. 2009. The movement analysis of temporal adverbial clauses. English Lan- guage and Linguistics 13: 385-408. Haegeman, Liliane. 2010. The Internal syntax of adverbial clauses. Lingua 120: 628-648. Hukari, Thomas E. and Robert D. Levine. 1995. Adjunct extraction. Journal of Linguistics 31: 195-226. Huddleston, Rodney and Geoffrey K. Pullum. 2002. The Cambridge Grammar of the English Language. Cambridge, UK: Cambridge University Press. Iatridou, Sabine. 2000. The Grammatical Ingredients of Counterfactuality. Linguistic Inquiry 31.2: 231-270. Iatridou, Sabine, and David Embick. 1994. Conditional Inversion. Proceedings of the North Eastern Linguistic Society 24: 133-147. Cambridge, MA: MIT Press. Karttunen, Lauri, and Stanley Peters. 1979. Conventional Implicature. In C.K. Oh and D. A. Dinneen (eds.), Syntax and Semantics 11 Presupposition: 1-56. NY: Academic Press. Kim, Jong-Bok, and Peter Sells. 2008. English Syntax: An Introduction. Stanford, CA: CSLI Publications. Larson, Richard. 1990. Extraction and multiple selection in PP. The Linguistic Review 7: 169-182. Newmeyer, Frederick J. 1998. Language Form and Language Function. Cambridge, MA: MIT Press. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. Longman.
