Quite in Acadian French : A variety of scalar Focus *

This paper discusses the properties of the quite de construction in the Baie SainteMarie, Nova Scotia variety of Acadian French in examples such as le/un quite de souper ‘quite the/a supper’ which have a hyperbolic connotation. Unlike epithet nouns and other French binominals (and Romance more generally), we argue that this Acadian French construction contains a complex DP with a unique N with an expanded left nominal periphery (cf. Giusti 2005, 2006, 2012). In order to unify its syntactic and semantic properties, we propose an analysis where [Adv quite] is a scalar Focus head (following Kayne’s 1998 treatment of scalar only), with a null M(easurement) Op(erator) (Rett 2008, 2011) in its specifier. This M-OP is associated with a degree argument (i.e. N) bound by existential closure, which forces its linearization within the predicate domain (cf. Diesing 1992). Lastly, we show that de in the AF ‘quite de’ construction cannot be construed as a linker (pace den Dikken 2004, 2006) or as a P category (pace Kayne 1994), but it is best analysed as a nominal functor (cf. Ihsane 2013) at the Comp-Infl interface whose role is to recategorize a referential DP into a property denoting DP.

Semantically, the presence of quite emphatically evaluates the lexical noun, on par with English.Syntactically, unlike in English, where quite is an adverbial modifier targeting a variety of XPs as in [quite [ AdjP expensive]], [quite [ AdvP frankly]], [quite [ DP the story]], [quite [ PP out of the question]], in Acadian French, quite is constrained to constructions as in (1) where it must follow the Determiner and must precede de.
To the best of our knowledge, this is the only variety of French with such a construction, arguably induced by contact with English. 2 Consequently, its syntax begs clarification, especially since the obligatory presence of de in (1) points to well-known complex nominal constructions (i.e.binominal constructions) available to most French dialects and to Romance more generally.
This paper discusses the properties of the quite de construction in relationship to binominal constructions and argues against such an analysis for the Acadian French variety.Rather, we show that, in order to unify its syntax and its 1 We use the following abbreviations in this paper: 'AF' Acadian French, 'FEM' feminine gender, 'MASC' masculine gender, 'SG' singular number, 'PL' plural number, 'DE' uncategorized de.Acadian French in general is marked by both conservatism, i.e. retention of vernacular features of French long lost from most spoken French varieties, and innovation, typically but not exclusively in terms of borrowing from English (see King 2000King , 2013 for discussion) for discussion).The phenomenon discussed here is expected to occur only in those Acadian regions of Atlantic Canada where there has been longstanding contact with English, such as Baie Sainte-Marie.For instance, while Emilie LeBlanc (p.c.) reports that part of the pattern described here is also found in southeast New Brunswick Acadian French, it is clearly not found in Acadian regions where there is less intense contact, such as northeast New Brunswick, or where contact has been of shorter duration, such as western Newfoundland.
Isogloss 2016, 2/1 3 semantics, quite in AF must be analysed as a scalar Focus head in an expanded left periphery of a uni-nominal, albeit complex, DP.
The paper is organized as follows: having concluded the introduction, section 2 discusses quite de in AF and compares it to binominal constructions in French more generally; section 3 focuses on some of its preliminary syntactic properties, such as agreement, modification and D type; section 4 engages with and dismisses previous analyses, while section 5 unifies the syntax and semantics of 'quite de' under the scalar Focus proposal and provides further support for the current analysis from exclamatives and other evaluatives.Section 6 concludes the paper.

Introducing quite de and binominal constructions
As King (2013: 102) points out, the data in (1) bear a strong resemblance to the use of French epithet nouns, exemplified by un espèce de cochon 'a real pig', literally, 'a sort of pig' and un putain de livre 'a bloody book', literally, 'a whore of book'.From a semantic point of view, both quite de constructions and epithet nouns are strongly evaluative, having a hyperbolic connotation.Crucially, quite has lost its original lexical meaning (< Anglo-Norman 'without opposition', c. 1225, OED), indicating instead something akin to "extreme degree", an issue we revisit in Section 4.
(3) a. partitive: C'est un meilleur morceau de terre.'It's a better piece of land' b. quantitative: Il y avait beaucoup de brume.'There was a lot of mist' c. possessive: la soeur à Marie 'Mary's sister' Unlike AF, other varieties of Canadian French (i.e.Laurentian French), however, do allow evaluative (Det) N1 de N2 constructions, as illustrated by: un criss de livre 'a bloody book' (literally, 'a Christ of book'), des hosties de questions 'some fucking questions' (literally, 'some hosts of questions'), from Drescher (2009).As both are semantically evaluative, it looks like AF has replaced the epithet N construction with the quite de construction.The question is whether the initial Adverb status of English quite is preserved in AF or whether we are dealing with a qualitative binominal construction, in which case quite now has nominal properties.Or perhaps something else altogether is going on.In order to tease these possibilities apart, one has to investigate the properties of the AF quite de construction.In the next section we start by looking at some preliminary properties.

Preliminary properties: Agreement, modification, and Det type
In this section, we explore phi-feature agreement, modification possibilities and determiner type associated with the complex nominal of the AF quite de construction.

Phi-agreement
Insofar as French qualitative binominals are concerned, Hulk & Tellier (2000) show that agreement is variable, in that it can be with either N1 or with N2 (i.e. the N preceding de or the N following de, respectively).Crucially, this choice is dependent on the amount of semantics associated with N1 (Doetjes & Rooryck 2001); consider (4), where the head/agreeing N is underlined: (4) a. Ton phénomène de fille est bien distraite.your masc phenomenon masc of daughter is quite absent-minded (fem) 'Your phenomenon of a daughter is quite absent-minded.' (Hulk & Tellier 2000: 36 Ce bijou d'église romane a été reconstruit(*e).that jewel masc of-church fem roman has been rebuilt masc / *fem 'That jewel of roman church was rebuilt.' (Doetjes & Rooryck 2001: 2(3b)) In (4a), phénomène 'phenomenon' has lost its original meaning, so fails to trigger agreement, while in (4b), bijou 'jewel' retains its semantics so it triggers agreement.Compare also the epithet use of espèce in un masc espèce fem de cochon masc 'a real pig' with its literal use in une fem espèce fem de poisson masc 'a species of fish': it is only in the latter case that espèce determines the gender of Isogloss 2016, 2/1 5 the entire DP phrase, as evidenced by the feminine indefinite determiner.In sum, the lexical heavy-weight between N1 and N2 determines the properties of the entire DP, so it becomes its head.
The AF data in (1a,b) show feminine gender agreement while those in (5a,b) show masculine agreement.(5c) shows plural agreement.
( (AF) 'It seems that it is going to be quite a/some wedding.' Crucially, the data in (1) and ( 5) show that quite does not interfere with the gender of the complex DP; rather it is always N2 that determines the phi-features (i.e.gender and number) of these evaluative DP constructions in AF.This is unsurprising as quite is semantically bleached and likely does not qualify as an N1 or we would expect to see AF constructions of the type in (2d) more productively.Modification facts discussed below equally rule out quite as an N category.Hulk & Tellier (2000) point out that French qualitative binominals allow modification on both N1 and N2, as seen in ( 6), where we also bracket the relevant constituents:

Modification
(6) cet [affreux jojo] de [mari de ma souer] 'this nasty sod of a husband of my sister' (Hulk & Tellier 2000: 34) However, unlike qualitative binominals in standard French, the data in (7) show that N2 but not quite may be modified by an adjective in AF, regardless of number properties.This confirms that quite in AF is not an N, but rather an adverb on par with English, and that the quite de construction in AF is not binominal, but rather contains a single N (i.e. the equivalent of N2 of otherwise legitimate binominal constructions in Romance).

Det type
So far we have seen that this complex DP allows for variable phi-fetaures.This subsection further shows that it allows for variable definiteness.In ( 1) and (5a-b), the DP containing quite de is indefinite, but in (5c) it is definite.( 8) further shows that both (in)definite options are palatable.
(8) a. du quite de souper (AF) of.themasc quite DE dinner masc b. le quite de souper (AF) the masc quite DE dinner masc Unsurprisingly the (in)definiteness options correlate with further semantic asymmetries.In particular, (9) shows that, in the appropriate contexts, the indefinite correlates with either an entity or an event reading while the definite forces an entity reading (Comeau, p.c.).( 9) a. Ça c'était un quite de souper.
(AF) 'That was quite a supper.' (entity or event reading) b.
(AF) 'That was quite the supper.' (speaking only of the food itself -i.e.entity reading) Having covered some of the preliminary properties of this complex DP in AF, we next survey two of the major previous analyses proposed for qualitative binominals in French/Romance and see what light they shed with respect to the quite de construction in AF.

Previous analyses and beyond
This section discusses the small clause analysis proposed for qualitative binominals in Romance (primarily Kayne 1994 andden Dikken 2006) as well as Doetjes & Rooryck's expanded DP analysis.Crucially, the main issue is to try and determine what the syntactic status of de is in these nominals.

Small Clauses: subject-predicate structures
For Kayne (1994), de in qualitatives, as in possessives, lexicalizes a D/P head and is comparable to a prepositional complementizer, as shown in ( 10): (Kayne 1994: 106 ( 108)) (10) also captures the fact that in these binominal constructions the structure contains a subject-predicate small clause in which the predicate undergoes inversion.
Isogloss 2016, 2/1 7 Based on their different semantics, a fact arguably first observed by Napoli (1989), Den Dikken (2006) splits qualitative binominal phrases into two types: (i) comparative, as in (11a), and (ii) attributive, as in (11b).The author points out that, in (11a), a comparison is drawn between a village and a jewel (i.e.'the village is like a jewel'), while for (11b), the more salient interpretation is one where idiot is a property ascribed to doctor, on par with attributive constructions like big for a butterfly, which are analysed as having a predicate-specifier structure.In particular, in attributive constructions, N1 (Milner's 1978 noms de qualité 'quality nouns') and N2 are base-generated in their surface word order, with the predicate nominal (i.e.idiot in (11b)) preceding its subject.For (11a), on the other hand, den Dikken argues for an underlying canonical predicate-complement structure, with predicate inversion (i.e. a village is base-generated in the specifier of the predicate phrase, with jewel, the nominal predicate, as its complement).Furthermore, in den Dikken's (2006) analysis, all instances of subject-predicate relationships are mediated by a 'relator' (with predicate inversion structures also having a 'linker').In French qualitative binominals, the relator is de (i.e. a nominal copula).In sum, Den Dikken argues that the non-equivalent semantics of these binominals points to two different syntactic structures (predicate-specifier, versus predicate-complement), with N1 being the predicate in both cases.
Importantly, predicate inversion (i.e. the comparative type) seems to trigger an emphatic reading, a situation mirrored in other Romance languages.For example, in Romanian, comparative clauses directly specifying degree, normally appear after the adjective, as in (12a); however, when preposed, they yield an emphatic reading and trigger de 'of' insertion, as in (12b): Given the emphatic semantics associated with the quite de construction in AF, such an analysis would be very tempting.However, the comparative binominal analysis is untenable on at least two grounds: (i) there is no comparison between quite and 'N2' (e.g.visite in (1a)), as there is between jewel and village in (11a) or bijou and voiture in (2d), and (ii) examples like (13) rule out a predicative analysis for quite.
(13) *La visite est (une) quite.the fem visit is a fem quite Perhaps unsurprisingly, this situation resembles that of epithet nouns, which also rule out a comparative reading.Doetjes & Rooryck (2001) specifically show that if N1 has lost its original lexical meaning, the binominal construction excludes a comparative reading.This is shown in ( 14). ( 14) a.
*The quality of your daughter is such that she resembles a phenomenon.b.
The quality of the church is such that it resembles a jewel.(Doetjes & Rooryck 2001: 2(6a-b)) Note that the quantitative construction with beaucoup/peu de also resists being paraphrased as a comparison, as seen in ( 15): (15) *The quantity of books is such that it resembles a lot.(Doetjes & Rooryck 2001: 2 (4b)) So, quite de constructions in AF resemble not only epithet Ns, but also quantitatives with beaucoup/peu (i.e.'pure degree' constructions, in the sense of Doetjes & Rooryck 2001).And, interestingly, as with quite de in AF, these quantitative constructions also force agreement with N2 rather than with N1; compare (16a)-(16b) -underlining our own: In sum, in this section we have shown that a subject-predicate analysis of the type proposed in Kayne (1994) and den Dikken (2006) is difficult to maintain for quite de in AF, but that this construction shares properties with both epithet nouns and the beaucoup/peu de construction.The next section explores the possibility that, rather than a verbal functor, de is in fact a nominal functor.20))

Expanded
The structure in ( 17) is appealing given the presence of an adverbial Evaluative head (à la Cinque 1999) and the potential of AF quite as an adverb.In particular, since in AF quite lacks both nominal and predicative properties and is derived from the English [ Adv quite], its status as an adverbial Evaluative head seems straightforward.Nonetheless, a structure of the type in ( 17) is likely incorrect given the en-cliticization properties of 'quite de' nominals in AF.Kayne (1975Kayne ( , 1994) ) argues that en-replacement is equivalent to a [ PP de NP] (i.e. a pro PP), given that it substitutes not only for complements to verbs, but also substitute for adjectives and nouns.Interestingly, while 'pure degree' binominals rule out en-cliticization, as seen for qualitative binominals in (18) and as expected under a structural analysis of the type in ( 17), the quite de construction in AF does not.
*Il en est un espèce de cochon.he of-it is a type of pig b.
(AF) 'She told quite the one.' b.
Il en a acheté une douzaine de pommes.'She bought a dozen (of them).' b.
Elle en a lu beaucoup des livres.'She read a lot of them.' (i.e.books) (21) Bien sûr que je m'en souviens de mon grand-père.'Of course I remember him.'(i.e.my grand-father) (22) J'en ai lu beaucoup/peu de livres.'I have read many/few.'(i.e.books) The en-cliticization property of our construction further rules out any structural similarities with qualitative/evaluative epithet binominals and dismisses Doetjes & Rooryck's (2001) analysis suggested in (17).If Kayne's PP equivalence analysis for en-replacement were correct, one possibility would be to assume that quite in AF modifies or is combined with a PP, where de is a preposition rather than a nominal functor.However, the PP analysis is problematic (pace Kayne) given that in partitives of the type in (20a) and the quantitatives in ( 22) no determiner occurs after de.Arguably, Ihsane's (2013) analysis, which proposes that, in addition to [ PP de NP], en may replace subparts of nominal structure (i.e.non-referential domains) is better equipped to capture all the French facts.
In the next section we explore some of Ihsane's (2013) tests with a view at trying to establish whether de in AF is best viewed as a P or as a functional nominal head (since we have established that it cannot be verbal in nature).Ihsane (2013) looks at extraction facts and presence versus absence of a determiner in order to decide on the status of de.In particular, since extraction out of a PP is generally not possible (arguably because these are phasal domains, Citko 2014 and references therein), in (23a), deux livres 'two books' and beaucoup de livres 'a lot of books' do not involve PPs since de Zola 'of Zola' can be felicitously extracted out of the 'de NP' domain.Conversely,in (23b) In addition, a determiner follows de in (23b) but not in (23a).Since P subcategorizes for determiners, de in quantitative constructions, as in (23a), cannot be a P. Rather, in such cases, it is a functional head in the nominal Inflectional domain.

On the status of de in AF 'quite de' constructions
Replicating these tests for the AF quite de construction, we get the following results.First, the data in ( 24) show that extraction is ruled out.
J'ai eu une quite de soupe aux choux.
(AF) 'I had quite the cabbage soup.' b.
*C'est aux choux que j'ai eu un quite de soupe aux choux.(AF) '*It's cabbage that I had quite the cabbage soup.' However, we argue that the ungrammaticality of (24b) is orthogonal to the PP facts.In particular, while extraction is clearly not possible, it does not immediately follow that de in these AF constructions is a P. If that were the case, Isogloss 2016, 2/1 11 one would expect co-occurrence with a determiner, but (25) shows that to be ruled out.
(AF) Intended: 'It seems that it is going to be quite the wedding.' b.
(AF) Intended: 'You had had quite the/a visit.'Both plural and singular definite determiners, in (25a) and ( 25b) respectively, are impossible with quite de in AF, while judgments are less categorical for un(e) 'a' in (25b).Following Hulk & Tellier (2001: 43) who show that in Dutch and English, een 'a' and a, are permissible with qualitative binominals (e.g. a hell of a man), but argue that this is not the indefinite article but an affective operator lacking phi-features, we assume that the same holds for the AF quite de construction.This, in effect, rules out P status of de here.
Having ruled out the P status of de based on the impossibility of a determiner, we need to find an alternate account for the facts in (24b).We propose instead an explanation based on focus intervention effects.Specifically, (24b) contains a cleft and French (including AF) uses clefts as a (contrastive) focusing strategy, as seen in ( 26).
(AF) 'It's to Mary that I gave the book.' (not to someone else) Since in Section 4 we argue that quite is involved in scalar focus in the AF quite de construction, (24b) is independently ruled out by the impossibility of having two simultaneous operator foci.
In sum, N2 must be a bare NP in quite de constructions, on par with qualitative binominals in French (Hulk & Tellier 2001) and beaucoup/peu de constructions.Furthermore, since only referential nominals (Chomsky 1981) -i.e.DP (but not NP) -need Case (Kayne 1999) -de cannot be assumed to be a caseassigner in these constructions in either standard or Acadian French (pace Jones 1996, Cardinaletti & Giusti 2006).Rather, Ihsane (2008: 168) argues that, in standard French, its function is to indicate N2 as a property/kind/category, rather than an object, on par with the Indo-European genitive it has replaced.While we do not take this to be its function in the AF quite de construction, we do follow Ihsane in concluding that it is a functional head in the nominal domain.

Summing up syntactic properties
Before summing up, there is one last syntactic property that is worthy of mention.In particular, as shown in ( 27), the AF quite de construction cannot occur in preverbal subject position in either the active or the passive voice.The constructions in (27) can, however, be salvaged via clefting, as seen in ( 28).Crucially what clefting does is reposition the quite de construction within the predicate domain.
(AF) 'It's quite the kids that have arrived.' b.
C'est une quite de visite qu'a été fait par mes amis.
(AF) 'It's quite the visit that my friends made.' In sum, the quite de in AF is constrained to a vP-internal position.Den Dikken (p.c.) points out that the English quite the construction is equally postverbally constrained, which is unsurprising given their similar extreme degree semantics.While the same linearization requirements are true of quantitatives like beaucoup/peu de (Ihsane 2013), we point out that this is likely due to their semantics of indefinite quantity (which also explains their obligatory cardinal reading), rather than their syntactic similarity to evaluatives (i.e.lack of a P de and subcategorization for a bare NP).
In sum, the properties discussed so far are represented in Table 1.Properties for quantitative and qualitative binominals hold for standard French.Undiscussed (and not immediately relevant properties) are represented via a dash (i.e.'-').  1 shows that the AF quite de construction differs syntactically from both types of French binominals under discussion.This warrants a novel syntactic representation for the AF quite de construction -an issue we expand on in the next section.

Unifying syntactic and semantic properties
This section unifies the syntax and semantics of quite de constructions in AF by proposing an account reliant on the presence of a scalar Focus projection in the left periphery of a complex nominal.This analysis is further supported by the behaviour of exclamatives and other AF evaluatives.

Quite as Focus in an expanded nominal left periphery
In her discussion of English quite, Diehl (2005: 15) points out that in expressions like quite a memory, quite is used as a focusing strategy to emphasize high degree linked to memory, while at the same time enhancing the positive evaluation that is implied. 4We have seen that the same is true for quite de in AF.
In order to accommodate its semantics, we propose that quite in the AF quite de construction instantiates scalar focus measuring extreme degree.According to Traugott (2006), scalar focus modifiers (e.g.even, only) involve the speaker's assessment and evaluation of intensity, position on a scale, ordering of alternates, and so on.Krifka (2007) also argues that scalar focus alternatives are ordered, with the focus denotation being the least or the greatest element. 5 In order to accommodate its syntax, we further capitalize on work by Giusti (2005Giusti ( , 2006Giusti ( , 2012) ) who proposes that DPs have left peripheries hosting TopP and FocP on par with clauses, with D equivalent to Force and 'd' equivalent to Fin, as seen in ( 29), which mimics Rizzi's (1997, et seq.) split CP.
(29) DP > TopP* > FocP > TopP* > dP (Giusti 2005: 35(28b)) However, the association of [ Adv quite] in AF with the domain of Focus in ( 29) still leaves open a number of questions.In particular, we have to determine whether this item resides in the head or specifier position of the Focus projection bearing the [scalar] feature.In addition, we need to fine-grain the category(c)al status of de in our AF construction: while section 4.3 establishes that it is not a P but rather a nominal functor, the introduction of the split left-periphery in (29) begs clarification as to whether it is an Infl or a C element.

Quite as a Focus [SCALAR] head
There is both conceptual and empirical support for treating the AF [ Adv quite] as a Focus head rather than an adverbial operator occupying Spec,FocP.
First, Kayne (1998) has already argued for treating only, another scalar focus item according to Traugott (2006), as a Focus head.Second, quite itself does not have a semantic value equivalent to "highest degree", so its status as an operator would not be semantically justified.However, since Focus heads are present in the derivation to accommodate an operator, the question arises as to what type of operator would be at stake in these AF constructions.We turn here to 4 See also Quirk et al (1985).5 Note that, in this respect, qualitative binominals are semantically distinct as these do not involve "a high degree" (Hulk & Tellier 2000).While the latter clearly express the speaker's strong evaluation of N2, unlike quite de in AF, they are not scalar (i.e. they do not involve a scale).Rett's (2008Rett's ( , 2011) ) null "M(easurement) operator" (henceforth, M-OP), essentially "a function from entities to degrees along some scale which may or may not be a quantity scale" (Rett 2011: 425).Basically, M-OP can occur freely with entities that can be measured and accounts for the fact that dimensions of measurement vary based on the noun involved and on context; we argue that this is what occupies the Spec,FocP of the AF quite de construction.
Empirical support for the above claims comes from at least the following two facts: (i) the inability of [ Adv quite] to combine with adverbial modification, which is suspicious if this were an XP in Spec,FocP, and (ii) linearization requirements of the construction more generally.The former is illustrated in (30), while the latter is discussed below.
(AF) it-is really a really quite DE dinner 'It was really quite a dinner.' The presence of M-OP in Spec,FocP can also account for the otherwise mysterious postverbal linearization demands of this construction.In particular, Rett (2011) argues that the degree argument introduced by M-OP is bound by existential closure.Since, following Diesing's Mapping Hypothesis (1992), clausal mapping reduces existential closure to the predicate domain (her VP, currently vP), as in (31), the postverbal requirement of the quite de DP shown in section 4.4 follows in a straightforward manner once we take into account that in Romance the verb raises to Infl (Emonds 1978).In sum, we conclude that the quite de construction in AF contains an expanded left-periphery with a Focus [SCALAR] domain hosting an M-OP in its specifier and quite in its head position.Lastly, in the next section, we revisit the status of de as an Infl or a C nominal element.

De as nominal Infl moved to d/Prop(erty)
In their discussion of standard French complex noun phrases of the type une pizza de chaude 'a hot pizza', den Dikken & Singhapreecha (2004) convincingly show that chaude 'hot' instantiates old information, while the entire DP receives a Isogloss 2016, 2/1 15 contrastive focus interpretation.The authors argue that these interpretive effects are the result of predicate inversion, which necessitates overt manifestation of the linker de.Similarly, in quite de constructions in AF, the constituent following de (i.e.N2) represents old information, while the entire DP yields a scalar focus interpretation.Furthermore, the presence of de is intrinsically linked to the presence of quite, (32): (AF) 'We had quite a good supper.'However, in Section 4, we ruled out a subject-predicate analysis for quite de in AF, so the presence of de cannot be due to predicate inversion and so de is not a linker.
Another possibility would be to assume that, akin to de/di in Romance infinitives (Kayne 1999, Rizzi 1997), de is the Fin complementizer category (NumP in Giusti 2012), forced to lexicalize whenever D and d project separately (i.e. in the presence of an expanded left-periphery).However, it is often the case that either Force ('D' here) or Fin ('d' here) remain null in split left peripheries (Rizzi 1997, Roberts 2001, inter alia) and the nominal left-periphery is already lexicalized twice in these AF constructions (i.e. the determiner and Focus quite), so there is no strong evidence in this direction either.
This takes us back to Ihsane's (2008Ihsane's ( , 2013) ) analysis of de as a nominal Infl element.Ihsane (2013) convincingly argues for [ F de] as an inflectional head positioned below Num(ber)P, the highest Infl head in Ritter (1991).Since in quite de constructions, Infl is not overtly realized in any other way, it is plausible to assume that de instantiates the Infl domain.Nonetheless, since de also has the merit of reanalysing/recategorizing a referential DP (e.g.la/une visite 'the/a visit') into a property denoting DP (i.e. de visite), it must be intrinsically linked to the Prop(erty) head, which Ihsane (2013) equates with 'd', essentially, a C type head.
In sum, we conclude that de merges as an Infl category that undergoes nominal I-to-C head movement to d/Prop.It has the dual function of both lexicalizing the nominal Infl domain and regategorizing the DP semantic type.This yields (33), which captures both syntactic and semantic properties of AF quite:

Exclamatives and additional data ('godam')
Before concluding, we think it worthwhile to point out that the Baie Sainte-Marie Acadian French variety has also borrowed God damn from English.The data in (34) show that godam de N also patterns like quite de in terms of its strongly evaluative nature and phi-feature agreement (i.e. it tracks N2).Furthermore, (34b) shows that modification is permitted on N2 but not on godam, while (34c) shows that en-cliticization is an option.