Finite Subjunctive ُ versus Non-Finite Infinitive Control Juxtaposed: A Cross-Linguistic Account
IntroductionEver since the onset of generative tradition in the nineteen fifties, probing into the nature of the theory of grammar received its greatest impact from the English-type accounts. Finite theory of control, as such, was not considered a mainstream approach at the time (Landau, 2004; Landa...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | Persian |
| Published: |
Alzahra University
2025-02-01
|
| Series: | زبان پژوهی |
| Subjects: | |
| Online Access: | https://zabanpazhuhi.alzahra.ac.ir/article_8365_b3645cf5d83146ae28ff8d95a905e291.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | IntroductionEver since the onset of generative tradition in the nineteen fifties, probing into the nature of the theory of grammar received its greatest impact from the English-type accounts. Finite theory of control, as such, was not considered a mainstream approach at the time (Landau, 2004; Landau, 2013; Landau, 2024) with PRO appearing only in nonfinite clauses, receiving no case (Chomsky, 1981). As a result, this study aims at investigating control constructions in Persian in their finite subjunctive clauses and comparing and contrasting them to the infinitive control constructions typically found in languages like English.To do so, we juxtapose the clausal architecture of finite subjunctive control as seen in Persian and the clausal architecture of non-finite infinitive control as found typically in a language like English. Whereas control constructions in a language like English appear in their bare infinitive clauses, with no overt manifestation of formal properties, the same type of constructions in Persian appear productively in non-bare finite clauses of the subjunctives. This dissimilarity in clausal architecture of English and Persian can be compared and contrasted in (1) and (2) respectively.Sara tæsmim=gereft ke æz talar birun be-r-e.Sara decision=took.3Sg. that from hall out Sbj-go-3Sg.[1]“Sarah decided to leave the hall.”Generally, these differences amount to saying that the two constructions in the two languages have two different types of embedded clause architectures, i.e. infinitive vs. finite complementation for the relevant embedded clause. The architecture, interestingly includes several properties that mark these clauses absolutely different from the infinitives. In what follows, we will demonstrate these differences by contrasting the two types of constructions in English and Persian as seen typically in examples (1) and (2) above.The juxtaposition of the two types of embedded constructions, as seen in the two types of languages above, provides us with two relevant types of architectures for the embedded clause cross-linguistically, i.e. the non-finite infinitive clause and the finite subjunctive complementation. While the embedded clause in (1) is a typical non-finite clause of the type bare infinitive construction found in a language like English, the parallel construction in (2) is a finite construction in Persian (Darzi, 2008; Karimi, 2008a; Ilkhanipour, 2014; Pirooz, 2010; 2016). As a result of the finite complementation, the embedded clause in (2) includes person and number endings, and as such introduces a TP projection in the embedded clause with [parson], [number], and [tense], similar to that of the matrix clause. As such, these embedded clauses are finite clauses (Cowper, 2016) as the verbal element obligatorily receives person and number ending completely similar to that of the matrix clause verbal element. This divorces the infinitival nature of the embedded control construction in these constructions, and as a result, integrates finite complementation into the embedded control constructions, as found similarly in other languages (Landau, 2004; 2013; 2024; Terzi, 1997). This is interesting when we also find non-finite control in the Persian with nominal infinitives, or gerunds in this language as well (Darzi, 2001; Pirooz, 2011; 2021).Interestingly, PRO actively appears in finite control constructions, similar to the non-finite, infinitive found typically in a language like English, contra the GB approach of the time (Chomsky, 1981) and as such, values/checks structural Nominative in the course of the embedded clause derivation, contra the Null Case approach (Martin, 2001), divorcing the PRO Theorem (Chomsky, 1981) in these finite environments (Landau, 2006; Sigurdsson, 2008; Pirooz, 2016). The construction of an embedded control construction in a sentence like (2) above has a TP, with the [person], [number], and [tense] actively present in the derivation, where the verbal element obligatorily receives person and number ending. As such, PRO base-generated within vP moves Spec;TP to check Nominative structurally (Pirooz, 2016). These appear in (2) above, repeated as (3) below with the relevant elements added to the derivation, shown in boldface for the convenience sake.Sara tæsmim=gereft …[TP PRONom. [vP PRO æz talar birun be-r-e]].Secondly, whereas the embedded clause in (1) typically appears in a bare infinitive construction, the embedded clause in (2) appears in the subjunctive construction, with the subjunctive clitic be- prefixed to the verbal element. This makes it possible to introduce a mood phrase (i.e. MoodP) into the derivation, headed by the subjunctive clitic be- prefixed ultimately to the beginning of the verbal element. These elements that appear in (2) above, are included in the derivation (4) below in boldface the for convenience sake.Sara tæmim=gereft … [MoodP [TP PRO [vP PRO æz talar birun be-r-e]]].Thirdly, whereas the embedded clause in a control construction in a language like English does not hold a CP projection, the embedded clause in (2) inadvertently includes a CP projection, with the complementizer ke “that” appearing as the head of the CP (Darzi, 2008; Hashemipour, 1989; Ilkhanipour, 2014; Karimi, 2008a; Pirooz, 2010; Pirooz, 2016). The CP projection and its relevant complementizer head ke in (2) above, are added to the derivation shown in in boldface in (5) below:Sara tæsmim=gereft [CP ke [MoodP [TP PRO [vP PRO æz talar birun be-r-e]]]].Therefore, the difference between the infinitive-type English complementation for a language like English and the finite-subjunctive complementation in embedded control constructions can be linearized below.Finite vs. Infinitive ComplementationInfinitive complementation (English)… [InfP PRO …]Finite complementation (Persian)… [CP ke [MoodP be- [TP PRONom. [vP [ PRO … be-v …]]]]]]These differences undoubtedly show a change of the idea of the control constructions from the early English-type accounts of non-finite control complementation, to the inclusion of finite and/or subjunctive complementation in the minimalist accounts. More specifically, the finite-subjunctive theory of control, as we see here, and as we see cross-linguistically, results in a shift of attention from merely considering finite and/or subjunctive control constructions as something marginal to the mainstream control constructions of the GB-type era, to embracing them as a new typology of control constructions with a new clausal architecture that shares its architectural properties with those ones in the matrix clause architecture (Landau, 2004; 2013; 2024). Conclusion In what follows, we will demonstrate how this theory of finiteness, as we illustrated here and as are found in the relevant literature of control complementations in other languages, resulted in the development and/or extensions in the theory of control, starting from the pre-GB era to the most recent minimalist accounts.Firstly, as the theory of generative grammar took its earliest explanations from the English-type languages, the earlier accounts of the theory of control and PRO dissociated finiteness from control constructions altogether, and related them to infinitive complementation found in a language like English. Therefore, a language like Persian at the time with the finite, non- infinitive control construction (Hashemipour, 1988; 1989) or having a subjunctive clause having merely a bare vP (Wurmbrand, 2001) to lack a CP projection (Ghomeshi, 2001) was considered exceptional, or marginal to the conception of the infinitive control, found in the mainstream-type languages like English. Secondly, since English does not typically utilize the subjunctive complementation, or alternatively, since the class of subjunctives in English does not generate control constructions, the earlier accounts of control in subjunctive constructions in the languages that have these constructions appeared marginal to the mainstream generativist accounts, though these studies have a history of their own, beginning in the 1980s and the 1990s (1998; Terzi, 1997; Hashemipour, 1988; 1989), among many others.Thirdly, the earliest accounts of infinitive-type control constructions considered the empty category PRO to be ungoverned with no Case, appearing only in the subject position of the infinitive clause (Chomsky, 1981). This was again to the exclusion of this empty category PRO in the finite clauses in the theory of syntax. Subsequently, at the time PRO was considered an exceptional nominal empty category, appearing only in the subject position of a non- finite infinitive construction, and not in that of a finite construction. However, with the development of the idea that PRO is a regular empty category, checking regular structural Case, and having regular morphological case similar to any other nominal elements elsewhere in the sentence (Sigurdsson, 2008; Landau, 2008), a regular Case/case was valued/checked for the PRO, and was received morphologically by PRO in languages like Persian that show PRO having regular Case/case (Pirooz, 2016).Finally, with the overt manifestation of the CP, and with the overt complementizer head ke “that” in these control constructions (Darzi, 2008; Ilkhanipour, 2014; Pirooz, 2016; Karimi, 2008a), Persian illustrates that the presence of a controlled CP is a possibility in control constructions. This makes the language similar to other languages that have this type of projection in their embedded control construction (Witkos, 2010; Sevdali, 2012), and dissimilar to the infinitive-type languages with no CP in their control environments. Nonetheless, as these control CPs in Persian are porous to a matrix probe, they allow for the ϕ-features of PRO to be controlled, i.e. valued/checked phase-externally, the theory of finite control in Persian provides challenges for the Phase Impenetrability of Condition. [1]. 3Sg.: third person singular, Sbj.: subjunctive, tæsmim=gereft: a complex-predicate control verb |
|---|---|
| ISSN: | 2008-8833 2538-1989 |