How cool is that! A new ‘construction’ and its theoretical challenges

Abstract

This paper deals with the controversial connection between sentence type and illocutionary force. I focus on wh-configurations of the form How cool is that! and demonstrate that this recent phenomenon presents some interesting challenges to approaches that are concerned with the illocutionary potential of sentence forms. In particular, the exclamation component of How cool is that! cannot be derived from features of exclamative syntax, but rather is a cumulative effect of exclamative intonation, the respective adjective, and the degree reading of how. However, the V-to-C-movement property results in a special pragmatics that proper exclamatives (i.e., without V-to-C movement: How cool that is!) lack. I claim that How cool is that! asks for affirmation or objection on the part of the addressee, and that this addressee-oriented component is signaled by a dedicated class of modal particles in the German version(s) of How cool is that! To analyze these addressee-oriented exclamations (A-EXCs), I adopt a syntactic approach to the connection between sentence types and their potential illocutionary uses and thus argue that this connection is by no means arbitrary and unrestricted. I show that both the special pragmatic function of How cool is that! and the corresponding distribution of modal particles in the German counterparts can be accounted for by referring to a compositional conception of sentence-type meanings, and I hence disagree with recent approaches that deal with How cool is that! and its counterparts in German as constructions (i.e., arbitrary form-function mappings) only.

1 Introduction

In linguistic typology, a distinction is often made between so-called ‘major’ and ‘minor’ sentence types. There is an overall consensus that declaratives (1a), interrogatives (1b), and imperatives (1c) belong to the class of major sentence types, while exclamatives, for instance, probably instantiate a minor type (2); see, for example, König and Siemund (2007, 2013).

(1)	a.	That building is tall.
	b.	Has that building been extended?
	c.	Go to the top of that building!

(2)	What a tall building that is!

In this paper, I will focus on a subclass of non-standard uses of interrogatives, namely on special uses of how-interrogatives. I will present the new observation that utterances like English How cool is that! have in fact a hybrid status: they function as both speaker-oriented exclamations (like [2]) and addressee-oriented questions (like [1b]). Drawing on cross-Germanic evidence from English and German, I will specify the pragmatic contribution of How cool is that! as addressee-oriented exclamation (A-EXC). The evidence discussed in this paper will show that this unique pragmatic function is indicated by a combination of both non-syntactic means and compositional syntax. Since A-EXCs are characterized by a complex interplay of intonational, syntactic, and lexical factors, a construction-oriented analysis that can assemble these diverse aspects under one ‘construction’ seems to be a natural account at first sight.

In a seminal paper on German sentence types, Reis (1999) has distinguished between ‘derivational approaches’ and ‘correspondence approaches’. Correspondence approaches, in contrast to derivational views, leave more room for non-syntactic factors (e.g., intonation, pragmatics, lexical choice) to determine the illocutionary force of a syntactic utterance. These accounts assign several aspects of different linguistic levels to a particular sentence type and thereby correspond to the general construction-based view on language, according to which “generalizations about linguistic structure are formulated in terms of […] conventionalized clusters of features (syntactic, prosodic, pragmatic, semantic, textual, etc.) that recur as further indivisible associations between form and meaning” (Fried 2015, 974).

In this paper, I will challenge recent construction-oriented approaches to utterances of the form How cool is that! In particular, I agree with such approaches that the exclamatory reading (i.e., the speaker is surprised/astonished that x is cool to an extreme degree) is yielded by an (idiosyncratic) set of non-syntactic factors. However, I will argue that morphosyntactic operations such as V-to-C movement still encode an illocutionary meaning component that sets these forms apart from their corresponding wh-exclamatives. Accordingly, the constructionist view on How cool is that! must be complemented by a compositional approach to sentence-type meaning to fully account for the special pragmatics of How cool is that! and its counterparts in other languages like German. Note also that neither cool nor the demonstrative that are obligatory components of this configuration—another argument against its idiomatic status. As I will show below, one can use any lexical items. However, in what follows I will mostly use How cool is that! for illustration purposes.

The paper is structured as follows. In Sect. 2, I will first discuss and then dismiss the idea that utterances of the form How cool is that! could be analyzed as a subtype of wh-exclamatives. In particular, after having looked at both the left periphery (Sect. 2.1) and the wh-constituent (Sect. 2.2) of wh-exclamatives and How cool is that!, I will conclude that there is no defining feature of wh-exclamatives that could explain the exclamation reading of configurations like How cool is that! Section 3 presents the main claim of the paper, namely that utterances like How cool is that! are addressee-oriented exclamations (A-EXCs) and thus feature a special pragmatics, which can be derived from both their non-syntactic features (like intonation, nature of the adjective; Sect. 3.1) and their syntactic makeup (Sect. 3.3). Crucial evidence supporting such an approach that combines insights from ‘derivational’ and ‘correspondence’ approaches will be taken from the German language (Sect. 3.2), where modal particles play an important role in performing the speech act associated with the syntactic form of A-EXCs. Section 4 summarizes and concludes the paper.

2 How cool is that! Rethinking the formal characteristics of exclamations

Intuitively, if speakers utter How cool is that!, then they express that something has exceeded a certain degree of coolness that they had expected. In other words, these configurations express speaker surprise due to a violation of the speaker’s expectation(s). At the level of illocutionary force, How cool is that! can thus be considered an exclamation, that is, an expressive speech act that conveys that a particular state-of-affairs has violated the speaker’s expectations. As is well known from the literature on exclamations, they can be conveyed by different syntactic forms. Consider the following English examples, which are taken from Rett (2011, 412):

(3)	a.	(Wow,) John bakes delicious desserts!
	b.	(My,) What delicious desserts John bakes!
	c.	(Boy,) Does John bake delicious desserts!
	d.	(My,) The delicious desserts John bakes!

Looking at the data in (3), one could claim that exclamation is a speech act that does not correspond to a particular sentence type but rather to specific uses of several sentence types. Given the idea of derivational approaches that sentence types should be categorized based on their morphosyntactic features (see Sect. 1), there is thus no such thing as a dedicated sentence type for exclamations; potentially, any sentence type can receive an exclamatory interpretation, but this is due to non-syntactic features such as intonation that are (according to most derivational approaches) not relevant for defining sentence types.

On the other hand, there are influential accounts that postulate the category of an exclamative sentence type. For instance, we know that exclamatives can have a dedicated syntactic signature, such as what a in What a delicious dessert John bakes! These elements can thus count as markers of exclamatives and their distribution can easily be accounted for when we postulate the category of exclamatives (see, e.g., Grimshaw 1979).

Our key example now raises interesting issues regarding these questions because although clearly functioning as an exclamation, it has the syntactic form of a wh-interrogative and thus adds another option to the inventory given in (3). In Sect. 3, I will elaborate on the claim that this additional option can actually be characterized as an utterance type sui generis, namely as ‘addressee-oriented exclamations’ (A-EXCs). To prevent any confusion right from the start and to improve the readability of the following sections, I will use the term A-EXCs already at this point of the paper to refer to configurations of the form How cool is that!¹

With this terminological choice in mind, let us now first turn to a discussion of the theoretical possibility of postulating structural differences in the left periphery of the clause that nevertheless might distinguish A-EXCs from their information-seeking wh-counterparts (Sect. 2.1). I will argue that there is no need to postulate extra functional structure for wh-exclamatives in the first place, and, based on this claim, I will thus conclude that left-peripheral syntax cannot help us to distinguish A-EXCs from regular wh-interrogatives. In Sect. 2.2, I will then turn to the internal structure of the wh-constituent and argue that the nature of both the wh-element and the adjective in A-EXCs cannot help us distinguishing between exclamations and other speech acts either. Section 2.3 summarizes the observations and concludes by briefly mentioning the main characteristics that can be observed in the domain of intonation.

2.1 How cool is that! and the left periphery of exclamations

Looking at the English inventory of exclamation speech acts in (3) above, let us first assume, for the sake of argument, that at least forms like (3b) can be categorized as a separate sentence type because they do not work as regular wh-questions due to their V-in situ configuration.² This is in accordance with most of the literature on English exclamatives (see Siemund 2015).

In particular, one could thus say that wh-exclamatives have a dedicated syntactic structure in English and can clearly be distinguished from wh-interrogatives (examples and judgments from Repp 2013, 66):

(4)	a.	[CP How long [TP this train journey could take]]!
	b.	[CP How long [C’ could [TP this train journey could take]]]?
	c.	*[CP How long [TP this train journey could take]]?
	d.	*[CP How long [C’ could [TP this train journey could take]]]!

Accordingly, as far as the English language is concerned, our key example How cool is that! clearly features interrogative rather than exclamative word order. In terms of word order, it thus cannot count as an exclamative and can be used for both conveying a request for information (5b) and performing the speech act of an exclamation (5a); ‘#’ indicates pragmatic deviance in a dialogue:

(5)	a. A:	How cool is that!	[intended: exclamation]
	B:	# I don’t know.
	b. A:	How cool is that (e.g., on a scale from 1 to 10)?	[intended: question]
	B:	To my mind, it’s very cool (I’d rate it a 9).

Both (5a) and (5b) have the syntactic form of an interrogative, but only (5b) is interpreted as a question speech act. Example (5a), by contrast, is interpreted as an exclamation. That is, the utterance expresses that the degree of coolness is not in accordance with the speaker’s expectations, and the speaker conveys that the propositional content (that a salient entity in the ongoing discourse is very cool) is true. This propositional part of the utterance is the reason why Speaker B’s response is odd because in the response given in (5a), Speaker B addresses an issue that is already backgrounded in the ongoing discourse. For many, this ‘backgrounding’ is a signature property of exclamatives and can technically be characterized as a factivity presupposition (see below).

In sum, there is a clear difference in meaning between the two readings paraphrased in (5). However, there is no difference in their overt syntactic structure (i.e., word order). One theoretical possibility to nevertheless distinguish between the two readings (question vs. exclamation) on syntactic grounds is thus to refer to ‘covert’ syntactic structure that has been claimed for the left periphery of exclamatives. One prominent approach in this regard is the theory by Zanuttini and Portner (2003), who refer to the backgrounding of descriptive content already mentioned above in their syntactic proposal.

More specifically, Zanuttini and Portner (2003), among others, have argued that exclamatives have a factive operator that interrogatives like (5b) lack. In their work, they focus on Paduan, which is spoken in the Italian city of Padua. They argue that this Romance variety provides direct evidence that exclamatives contain an extra CP layer of structure. Without going into too much detail here, let me briefly illustrate their analysis for an wh-exclamative like (6a); see Zanuttini and Portner (2003, 64):

(6)	a.	Che	alto	che	l	ze!
		what	tall	that	subject.clitic	is
		‘How tall he is!’
	b.	[CP che alto [[C ∅] [CP FACT [C che] IP]]]

The main point of the data and the analysis in (6b) is that two CP specifier positions are needed in order to provide a phrase-structural position for both the factive operator (also needed for embedded contexts introduced by che ‘that’) and the wh-phrase che alto (‘what tall’). Zanuttini and Portner (2003) then argue, based on evidence showing that exclamatives always carry a factivity presupposition, that their analysis should be postulated for other languages too. Since exclamatives, according to their claim, are the only unembedded configurations that feature such a factivity presupposition, clause typing is taken care of by the factive operator, and there is no need to postulate additional projections like ForceP, let alone more specific ones like ‘ExclamativeForceP’ or the like. For such a Split-Force approach to exclamatives and further configurations that convey surprise on the part of the speaker, see, for example, Munaro and Obenauer (1999) and Obenauer (2006).

While approaches differ in their claims as to how much functional structure is needed to account for the syntax of exclamatives, many accounts in the spirit of Zanuttini and Portner (2003) share the central assumption that for languages like English, we have to postulate covert syntactic structure to syntactically distinguish between interrogatives and exclamatives.

In what follows, I would like to adopt a different approach and briefly illustrate the basic idea. Let us look at the exclamative counterpart of our key example How cool is that! in (7):

(7)	How cool that is!

My claim is that even the overt syntactic structure (lacking V-to-C movement) is not a syntactic feature that can be used to distinguish exclamatives from other sentence types. Note that the word order of the wh-exclamative in (7), despite the presence of a wh-phrase, is in fact the syntax of a declarative, featuring no subject inversion (see Ginzburg and Sag 2000, 224 on this parallel):

(7’)	[CP how cool [C ∅] [IP that [I is]…]]

Accordingly, only looking at the overt structure (word order), we see that exclamatives feature the syntax of declarative assertions. The only remaining structural difference in this domain is thus based on covert structure, namely the presence of the factive operator in a Split-CP approach like the one exemplified in (6b) above. But what if the evidence for such a factive operator is weakened?

It would take me too far afield to discuss all the semantic and pragmatic details that can be found in the literature on exclamatives to support the idea that exclamatives do not feature assertive force, but rather convey their descriptive content via factivity presuppositions. However, two facts are always mentioned, and they are mentioned as early as Grimshaw (1979) to support this idea: (i) exclamatives cannot be used as answers to information-seeking questions and (ii) the descriptive content of exclamatives cannot be denied. Let me briefly comment on these two points and start with (ii).

Fact (ii) has always been based on introspective judgments like the ones given in (8) and (9); see Rett (2011, 414):

(8)	A:	(Wow,) John bakes delicious desserts!
	B:	No (he doesn’t), these are store-bought. John’s actually a terrible cook.
(9)	A:	(My,) What delicious desserts John bakes!
	B:	? No (he doesn’t), these are store-bought. John’s actually a terrible cook.

In the literature, it has been claimed that the descriptive content of exclamatives (9), in contrast to declarative assertions (8), cannot be denied. However, this is not supported by recent experimental work on Spanish (Villalba 2017) and German exclamatives (Trotzke 2019), showing that the descriptive content of wh-exclamatives is indeed ‘at issue’ and can thus be denied. I refer the reader to these papers to see that judgments like the one given in (9B) do not stand up to empirical investigations that actually test and measure the acceptability of those patterns.

Fact (i) concerns Grimshaw’s (1979) claim that exclamatives cannot be used as answers to questions. The patterns we see in (10), the argument goes, suggest that while declarative assertions can serve as answers to information-seeking questions (10B), exclamatives are infelicitous, even if they express exactly the same descriptive content as the declarative counterparts (10B’). One thus concludes that this is so because the descriptive content of exclamatives is a factivity presupposition, hence not asserted; and that the pattern in (10) can be identified with the more general observation that a question cannot be answered by a sentence that presupposes the answer (11); I give below Grimshaw’s (1979, 321) data and judgments:

(10)	A:	How tall is John?	(11)	A:	Did John leave?
	B:	John is very tall.		B:	#It’s odd that he did.
	B’:	#How tall John is!		B’	#I’d forgotten that he did.

Again, my response to the above argument is that it is not valid. To understand what is going on in (10), I would like to suggest that (12B) below is a felicitous discourse move because it matches the constituent question in (12A). (12B’), on the other hand, causes a mismatch at the level of information structure. To see this, look at the following examples:

(12)	A:	How fast was Eliud Kipchoge*?	*current marathon world record holder
	B:	Eliud Kipchoge was [very]F fast.
	B’:	# [How fast Eliud Kipchoge was!]F

While (12B) is thus a perfect answer to a narrow-focus question (12A), (12B’) is odd, as already pointed out in the literature cited above. However, when the information-structural context is changed by using the two different utterance forms as responses to a broad-focus question, we see a clear improvement of the exclamative (13B’):

(13)	A:	Tell me, how did Eliud Kipchoge do in the race?
	B:	My god! [He was very fast!]F
	B’:	My god! [How fast he was!]F

I claim that (13B) is also fine in such a context because it no longer expresses narrow focus as in (12B), but instead (due to different intonation) it can also be interpreted as an all-focus declarative.

Given that exclamatives can be used as responses to information-seeking questions and thus feature assertive force, I hypothesize that a covert factive operator is too strong a claim and that exclamatives should better be analyzed as configurations featuring (i) declarative word order (lacking subject inversion) and (ii) conveying an all-focus interpretation. According to this perspective, exclamatives fall into one category with other declarative configurations that express surprise on the part of the speaker. In the recent syntactic literature, these configurations have been analyzed as ‘mirative focus’ (Cruschina 2012, 2019; Authier and Haegeman 2019) or cases of ‘emphasis for intensity’ (Trotzke 2017a; Beltrama and Trotzke 2019), and some recent experimental work shows that their interpretation can indeed be distinguished from other information-structural readings (Trotzke 2017b; Destruel et al. 2019).

In terms of the syntactic derivation, this would mean that [wh + Adj] in wh-exclamatives targets the left periphery to express surprise on the part of the speaker just like other phrases can be moved there to instantiate mirative fronting in declaratives (see literature cited above; e.g., Trotzke 2017b; Cruschina 2019). Let us adopt the syntactic implementation from Trotzke (2017a) and assume that the respective operator Emp_[intensity] is hosted in a functional projection below the question/assertion operators that are, according to cartography (e.g., Rizzi 2014), located in the domain of ForceP. Emp_[intensity] takes scope over the propositional core of the clause (FinP) (all-sentence focus, see above)³:

Crucially, data first mentioned by Bianchi and Cruschina (2016) suggest that the projection encoding the surprise of the speaker (here: Emp_[intensity]) must indeed be located below Force because otherwise German data like (15) cannot be accounted for:

(15)	a.	Du	fährst	morgen	ans	Meer?
		you	go	tomorrow	to.the	seaside
		‘Are you going to the seaside tomorrow?’
	b.	Ans	MEER	fährst	du	morgen?
		to.the	seaside	go	you	tomorrow

The emphatic effect of fronting in the polar question (15b) is interpreted with reference to the focus alternatives of the proposition p, and not with respect to its complement ¬p. Given that the polar set is generated by a Polar Question operator which takes the proposition p as input, it makes sense to postulate that the Emp operator is below the Question operator because Emp must be located at a compositional level where the complement proposition ¬p is not yet available.

The account by Trotzke (2017a) and Bianchi and Cruschina (2016) thus diverges from other more recent accounts that analyze the exclamation component as a modifier of illocutionary forces (e.g., Taniguchi 2017, 141–151). Although the approach presented here shares the basic assumption with Taniguchi (2017) that exclamatives do not have a force of their own, it conceptualizes the exclamation component along the lines of so-called mirative focus and thus locates the emphasis operator in the information-structural layer of the clause (see also fn. 3). According to this approach, exclamatives are nothing special and just another type of emphatic assertion, similar to declaratives with exclamative intonation, with lexical elements to mark unlikelihood like English even, or instantiations of mirative focus fronting.

Let us take stock. This section has provided some evidence for showing that there are good reasons to question the postulation of covert syntactic structure to distinguish between questions (16) on the one hand and exclamations on the other hand (17). In particular, both recent empirical work and recent theoretical proposals provide alternative ways to think about wh-exclamatives (17b), namely that they might fall into one class with other configurations displaying the same overt word order (declaratives lacking subject inversion).

(16)

How cool is that?

[question]

(17)	a.	How cool is that!	[exclamation]
	b.	How cool that is!	[exclamation]

Accordingly, the burden of proof is on those who propose that there is covert structure that would allow putting (17a) and (17b), which feature different overt structure (word order), into one syntactic class. In what follows, I will thus assume that the exclamation component of (17a) cannot be explained by a dedicated left-peripheral syntax. But what about the ‘interrogative’ component? In other words, the more interesting question maybe is how the inversion syntax (17a) contributes to the exclamation reading of A-EXC configurations like How cool is that! But before turning to this question in more detail, let us first discuss a further aspect of A-EXCs that might distinguish this utterance from other sentence types: the nature of the wh-constituent.

2.2 How cool is that! and the nature of the wh-constituent

After having looked at (and dismissing) the theoretical possibility of deriving the exclamation reading of A-EXCs from covert structure that is shared with wh-exclamatives, let us now turn to a second domain where (5a) and (5b), repeated here for convenience, could differ in terms of structure: the internal structure of the wh-phrase.

(18)	a. A:	How cool is that!	[intended: exclamation]
	B:	# I don’t know.
	b. A:	How cool is that (e.g., on a scale from 1 to 10)?	[intended: question]
	B:	To my mind, it’s very cool (I’d rate it a 9).

An interrogative like (18b) presupposes that something is cool to a certain extent. The corresponding exclamation (18a) carries the same presupposition. However, (18a) expresses that the value of the open variable is located at an upper point of a given scale, whereas the interrogative expresses that the variable is to be filled by the addressee. Despite an abstract functional similarity, we thus see that the semantic role of the lexical item how within exclamative clauses is different from that within wh-interrogatives; in exclamatives, it corresponds to ‘so’, the typical degree element.

This is a well-known property of exclamations: they can feature so-called degree uses of certain wh-pronouns. To illustrate, let us adopt an approach by Krifka (2011) to question speech acts and assume that while the degree of the respective property is understood to be extraordinary in exclamatives, it is simply unspecified in interrogatives. That is, similar to regular questions, A-EXCs like How cool is that! denote a set of alternatives (however specified, e.g., by a function or by a proposition set). But in addition, these alternatives, according to Krifka (2011), are ordered along a d(egree) scale. In a proposition-set analysis, this order can be expressed as follows (see Krifka 2011, 1775):

(19)	{${\langle}$λi[cooli(that ‘x’)(d)], λi[cooli(that ‘x’)(d′)]${\rangle}$ | d < d′)}

Utterances of the form How cool is that! are thus (just like exclamatives) formed over a variable that ranges over degrees, like ‘coolness’ in our example.

Sometimes the degree use of wh-pronouns formalized in (19) has even been grammaticalized. For instance, what in interrogatives functions as the head of a nominal phrase (What did you read?), or as a determiner (What book did you read?). In exclamatives, however, what functions as a predeterminer (20a) and, in this use, cannot occur in interrogatives (20b); see Elliott (1974, 235)⁴:

(20)	a.	What a fast car he is driving!
	b.	* What a fast car does John drive?

Such a distinction in terms of phrase structure also holds for other languages. It has been proposed that the degree reading of what, but also of how, has head status in exclamatives. Brandner (2010) has argued along these lines. That is, in German cases that are ambiguous between an interrogative and an exclamation interpretation (21a), she claims that wie (‘how’) is a head-element in the exclamation interpretation and thus behaves like was (‘what’) in (21b); examples from Brandner (2010, 108):

(21)	a.	Wie	bist	du	groß	geworden!/?
		how	are	you	big	become
	b.	Was	bist	du	groß	geworden!/*?
		what	are	you	big	become
		‘How tall you became!’

She provides a body of empirical evidence for the claim that both was and wie correspond to the degree element ‘so’, and that these elements adjoin in a clitic-like fashion to the finite verb in the C position⁵:

(22)	[CP	[C’	was/wie	bist	[IP	du	groß	geworden]]]
			what/how	are		you	big	become
	‘How tall you became!’

But can the non-degree and the degree use of wh-pronouns (and the corresponding phrase-structural differences claimed in the literature) now help us in distinguishing between the question and the A-EXC reading of How cool is that? Only looking at the phrase-structural status, I claim that these differences cannot account for the ambiguity of the string How cool is that. In particular, it is clear that how + adjective sequences (because of the modifying function of how) clearly have phrasal status and thus correspond to the general phrasal makeup we see in a regular wh-question, namely that the wh-element, together with the adjective, occupies the Spec-CP- and not the C-position in such questions:

(23)	[CP [AP how cool] [C’ is [IP that]]]

At this level of consideration, we can thus conclude that wh-questions (18b) and A-EXCs like How cool is that (18a) do not display differences at the structural level: both exhibit T-to-C movement and Spec-CP is filled by a phrasal constituent containing a wh-element.

Of course, there are structural differences when looking inside the wh-constituent—and these differences have been analyzed extensively in the literature (some of which has been cited above). As already mentioned above, the wh-pronoun in the exclamation reading of How cool is that! has a degree reading that corresponds to the degree marker so and can be formalized as follows (‘≫ θ’ = ‘going beyond a threshold’); see, among others, Kennedy and McNally (2005) for detailed discussion of the relevant degree semantics:

(24)	[[how cool]] = λx.cool(x) ≫ θcool

It is now interesting to see that the [wh + Adj]-constituent with the reading in (24) features a root/non-root asymmetry that has also been observed in the licensing of other emphatic wh-constituents that are often used in surprise contexts. In particular, as first noted by Lasnik and Saito (1984), English appears to disallow a wh-word with the hell to remain in situ (e.g., *Who is in love with who the hell?; see Pesetsky 1987; Den Dikken and Giannakidou 2002). The same holds for a [wh + Adj]-constituent such as (24) above. While wh-in situ in the degree reading sketched in (24) is not possible at all (25), it is perfect in the question reading (26); examples and judgments from Nye (2009, 19):

(25)	a.	How cool is that!
	b.	* That’s how cool!

(26)	a.	How cool is that?
	b.	That’s how cool?

We thus see that [wh + Adj] in its degree reading must access the clausal left periphery, and I would like to adopt an approach that has been proposed by Bayer and Trotzke (2015) for other constructions of the wh-the-hell type: the wh-constituent in (24) carries a root-sensitive feature of emphasis for intensity and therefore must be preposed to the relevant left peripheral projection of emphasis (see Sect. 2.1 above). Crucially, this projection—where, among other things, mirativity is encoded—cannot be at the level of illocutionary force. As already pointed out above, potentially any clause type can feature mirative interpretations, and so the relevant phrasal constituents are moved to the specifier of EmpP before they must move further for reasons of clausal typing.⁶ The difference between (25) and (26) is therefore that (26) lacks the extra movement to EmpP, and the [wh + Adj]-constituent in that case merely moves to the left periphery to form a (non-echoic) question.

To sum up, the different interpretations (and thus also the different phrasal compositions inside the wh-phrases) are due to the left-peripheral options that are not shared between A-EXCs and wh-exclamatives on the one hand and interrogatives on the other hand, but rather are shared between any emphatic-fronting configuration (including declaratives) on the one hand and any non-emphatic speech act interpretations (lacking the intensifying effect of scalar readings) on the other hand. All in all, the differences of the wh-constituents in the question and the exclamation reading of strings like How cool is that can be accounted for by referring to a general strategy of emphatization in the left periphery—and not by referring to a distinct feature of wh-exclamatives shared with A-EXCs.

2.3 Interim summary

We can thus draw the preliminary conclusion that the exclamation reading of A-EXCs cannot be formally distinguished from their wh-interrogative counterparts by referring to syntactic properties that have been claimed to exclusively hold for wh-exclamatives. More specifically, the discussion above suggests that A-EXCs like How cool is that! cannot be analyzed as a subtype of wh-exclamatives because (i) they exhibit a different word order and do not share covert structure that distinguishes exclamatives from other sentence types (Sect. 2.1), and (ii) although their wh-constituent differs from the one in wh-interrogatives, both the degree semantics of the wh-constituent and the corresponding left-peripheral movement are features that are shared with emphatic readings of speech acts more generally (Sect. 2.2).

Given the discussion above, the general picture that emerges is that the exclamation component of A-EXCs cannot be considered (and analyzed) by features that have been proposed for wh-exclamatives. In the next section, I will therefore turn to the following two questions:

(Qi)	Are there other non-syntactic cues (also in other languages) for the exclamation reading?, and
(Qii)	If the syntax of A-EXCs is ‘non-exclamative’, what kind of role does it play in the interpretation of the speech act?

I will thus turn to the question of how the inversion syntax of A-EXCs contributes to their reading. I will approach both questions by comparing English to German. German is very useful for comparison because it has almost word-for-word equivalents to English configurations of the form How cool is that! Consider the examples in (27), the German counterparts of our key example How cool is that:

(27)	a.	Wie	geil	ist	das	denn!	[intended: exclamation]
		how	cool (lit. ‘horny’)	is	that	part
		‘How cool is that!’
	b.	Wie	geil	ist	das	denn?	[intended: question]
		how	cool (lit. ‘horny’)	is	that	part
		‘How cool is that (e.g., on a scale from 1 to 10)?’

Although cases like (27a) are also a quite recent phenomenon in German, there are already two studies on this type of utterances (Finkbeiner 2015 and Auer 2016); both are framed within a construction-grammar account, highlighting that many linguistic dimensions of the utterance (27) accumulate to yield the exclamation reading: nature of the adjective, intonation, etc. Given that we have argued above that features of exclamative syntax cannot explain the exclamation reading of A-EXCs, it might be useful to look into this kind of literature to explore whether we observe reliable (non-syntactic) cues for the exclamation reading (Qi above).

Qii can also profitably be addressed by comparing German and English because in German both T-to-C movement and verb-final wh-configurations lacking this movement can express the speech act of an exclamation (i.e., there are both verb-final AND V2 wh-exclamatives; see Sect. 3.2 below). Accordingly, the same syntactic option used to convey an exclamation (V2 wh-exclamatives) can also be used to perform the speech act of a regular wh-question. If the inversion syntax of A-EXCs in English has any semantic impact, and if the German counterpart in (27b) really expresses the same, then one would expect that German conveys this semantic impact by other means than word order—and this is exactly what we find: see Sect. 3.2 below.

3 Connecting syntax to force: the special pragmatics of A-EXCs

3.1 Non-syntactic cues for the exclamation reading: English and German

Since we have argued in Sect. 2.1 above that features of exclamative syntax cannot explain the exclamation reading of A-EXCs, it might be useful to look into some of the literature that has investigated some non-syntactic cues for the exclamation reading. Let us first look at the claim that the nature of the adjective might be a good candidate for disambiguating between the question and the exclamation reading of configurations like How cool is that.

In particular, Finkbeiner (2015) has argued for the German version of How cool is that! (Wie geil ist das denn!, see above) that the use of evaluative adjectives like cool or ingenious result in ambiguity, whereas a non-evaluative adjective such as basal usually results in an interpretation in terms of a regular wh-question (Finkbeiner 2015, 260). However, I think the disambiguating function of the adjective is overstated in Finkbeiner’s (2015) study. While she might be right concerning the relative frequency of such adjectives, I hypothesize that any adjective compatible with a gradable reading can occur in A-EXCs, even those where a comparative form does not exist.

A corpus search in the Corpus of Web Based Global English (GloWbE; Davies 2013) indicates that adjectives such as wrong can occur in the exclamation reading without any problems. The GloWbE corpus is a resource that is particularly useful for searching expressions associated with informal, colloquial registers. The same holds for the German version of A-EXCs; see a parallel example from the DeReKo corpus in (29)⁷:

(28)	I bought Saving Valor from Amber Kell (definitely DRM free) and I can read it on the Kobo app on my iPad, but I cannot download it from the site to my computer. How wrong is that? I haven’t purchased from them since.
	<http://rabidreads.ca/2012/10/axe-to-grind-kobos-website.html>

(29)

Ich

finde

es

eine

Frechheit,

Änderungen

unreflektiert

(und

vor

allem

auf

I

find

it

a

cheek

changes

unreflecting

and

above

all

at

einmal!!!)

rückgängig

zu

machen.

So

macht

ihr

euch

alle

lächerlich.

Nun

once

undone

to

make

so

make

you

yourself

all

ridiculous

Now

steht

da

wieder

„Einige

Religionen[,]

die

aus

dem

Sufismus

stand

there

again

some

religions

that

out.of

the

Sufism

hervorgingen“(sinngemäß).

Wie

falsch

ist

das

denn?

originate.from OWTTE

how

wrong

is

that

part

‘How cheeky to undo changes without thinking (and what’s more, at once!). This way, you all embarrass yourself. Now, it says again “Some religions that originate from Sufism” (OWTTE). How wrong is that!’

[DeReKo: attested at http://de.wikipedia.org/wiki/Diskussion:Dreadlocks:Wikipedia 2011 on March 1, 2019]

Given this observation, we can conclude that although in many cases the nature of the adjective might help in disambiguating between the two readings, this component of the wh-constituent is not a one-hundred percent reliable cue.

Another non-syntactic cue that suggests itself is intonation. Nye (2009) has pointed out that the main stress in the exclamation reading of How cool is that usually falls on the subject (30a; without evoking semantic contrast), whereas it is usually on the adjective in wh-questions (30b):

(30)	a.	How tall is THAT  => !
	b.	How TALL is that  => ?
	c.	HOW tall is that  => ?
	d.	How tall IS that   => ?/!

Nye (2009) does not provide intuitions about the other options in (30c) and (30d), but a small group of English native speakers (n = 5) I have consulted uniformly tells me that (30c) cannot really be interpreted as an exclamation, while (30d) could receive both interpretations (i.e., exclamation interpretation and regular wh-question). However, it is interesting to note that the heavy stress on the subject (here: a demonstrative) is not an obligatory component of the exclamation reading. In her empirical appendix, Nye (2009, 60–69) reports cases where both the personal pronoun (31a) and the expletive (31b) it can appear as a subject in such exclamations, although it cannot bear heavy stress:

(31)	a.	How big is it!
	b.	How sweaty is it today!
	c.	How big IS it!
	d.	How sweaty IS it today!

Given what I have said above, the main accent would therefore most probably fall on the auxiliary when the speaker intends the exclamation reading in these cases (31c/d).

Be that as it may, as soon as a d-pronoun is present in those configurations, we often observe that it is heavily stressed, which is usually not the case in the wh-question interpretation. This intonational difference is also confirmed empirically by Repp’s (2015) recent work on comparing the acoustics of wh-questions and verb-second wh-exclamatives in German.

Although not concerned with wh-configurations that contain wh + Adj in Spec-CP, this study clearly shows that d-pronouns are more prominently stressed in wh-exclamatives, and that this stress pattern, although not obligatory, can thus serve as a cue for the exclamation interpretation in many cases. Note that just like we have seen for English in (30) above, there is also a preferred stress pattern for the exclamation interpretation of A-EXCs in German. In particular, the version where the strong (subject) d-pronoun is accented is the variant that most clearly signals the exclamation reading (32b):

(32)	a.	Wie GEIL ist das denn =>  ?/!
	b.	Wie geil ist DAS denn =>  !
	c.	WIE geil ist das denn   =>  ?
	d.	Wie geil IST das denn =>  ?/!

Supporting this idea, Auer (2016, 73) claims that in the exclamation interpretation, the main accent in fact must be placed on the demonstrative. This intuition can be related to the recent experimental work by Repp (2015) above, who shows that d-pronouns are more prominently stressed in wh-exclamatives, and that this stress pattern, although not obligatory can thus serve as a cue for the exclamation interpretation in many cases. However, as indicated in (32a), the exclamation interpretation can also arise in cases where the adjective is heavily stressed. This is in line with older (non-experimental) claims by Altmann (1993) that the so-called exclamative accent in German wh-exclamatives often occurs on the scalar expression. Altmann (1993) also claims that the exclamative accent may also be carried by the verb in V2 exclamatives, similar to what is indicated in (32d).

Given all these different hypotheses, the only two stress patterns that might display a preference for one of the two readings are (32b = exclamation reading) and (32c = regular-wh-question reading). This is also in line with what has been found in the production studies by Repp (2015): stressing the d-pronoun (and not stressing of the verb in second position) is the most reliable cue for the exclamation interpretation, while stressing the wh-element most of the time is associated with the question interpretation.⁸ We can thus conclude that the exclamation component of A-EXCs can indeed be signaled by both intonation and the respective nature of the adjective, but that these non-syntactic cues for this particular reading are not one-hundred percent reliable cues taken by themselves. However, taken together they can ‘add up’ and reliably result in the exclamation interpretation. This might be the reason why the German versions of A-EXCs have so far only been investigated from a constructionist perspective; this framework lends itself best to capture those cumulative effects coming from different linguistic levels/domains. In what follows, I will look at another non-syntactic means of A-EXCs in German that, according to my approach, plays an important role in the interpretation of the utterance: the distribution of modal particles (Sect. 3.2). Based on this discussion, Sect. 3.3 will then turn to the comparison with English again, and I argue that the word order in English A-EXCs actually impacts their reading, and that this impact can be analyzed along the lines of what I will now say about the modal particles in the next section.

3.2 German A-EXCs and the distribution of modal particles

As already mentioned at the end of Sect. 2 above, in German both T-to-C movement (33a) and verb-final wh-configurations lacking this movement (33b) can express the speech act of an exclamation. Accordingly, the same syntactic option used to convey an exclamation can also be used to perform the speech act of a regular wh-question (34):

(33)	a.	Wie	vergesslich	ist	Oma	aber auch	seit	letztem	Jahr	geworden	ist!
		how	forgetful	is	grandma	part	since	last	year	gotten	is
	b.	Wie vergesslich Oma aber auch seit letztem Jahr geworden ist!

(34)	Wie	vergesslich	ist	Oma	denn/wohl	seit	letztem	Jahr	geworden	ist?
	how	forgetful	is	grandma	part/part	since	last	year	gotten	is
	‘How forgetful has grandma gotten since last year?’

In our context it is important to note that the respective readings can also be indicated by the non-obligatory exclamative (33) and question particles (34), which only work in their respective speech acts. Given that German features particles like that, let us now focus on the obvious difference between English How cool is that and German Wie geil ist das denn: the presence of the modal particle denn (lit. ‘then’); on German modal particles in general, see, for example, Meibauer (1994); Grosz (to appear); and many others.

Finkbeiner (2015) points out that the modal particle can also be left out when the demonstrative is heavily stressed and thereby evokes the exclamatory reading. However, although not obligatory, the presence of a modal particle is highly preferred in the German versions of A-EXCs. Based on a corpus search in the DeReKo corpus, Auer (2016) found 565 hits of the form Wie ADJ KOP DEM-PRON (denn) and reports the ratio of 7:1 for the presence of the modal particle in the A-EXC interpretation. It is thus reasonable to look in more detail at what kind of modal particles can occur in German A-EXCs because their distribution might tell us something about the exact discourse function of these utterances.

Let us first look at Table 1, which is based on Thurmair’s (2013) overview of modal particles in different sentence types and summarizes the distribution of particles occurring in wh-exclamatives and wh-questions in German: In what follows, I will argue that the set of particles that can occur in German A-EXCs does not correspond to an intersection between particles that are licensed in wh-questions and particles that occur in exclamatives. Rather, the special pragmatic function of A-EXCs is signaled by a dedicated class of modal particles.

To see this, let us start at a very general level. As a first approximation, German modal particles tell us something about our general inventory of basic (or ‘major’) sentence types. It is often claimed that sentence-type sensitivity is a defining property of modal particles (e.g., Thurmair 2013). To illustrate, consider the following contrasts:

(35)	a.	Wo	fährt	er	denn	hin?
		where	travels	he	part	to
		‘Where is he traveling to?’
	b.	* Fahr	denn	nach	Barcelona!
		travel	part	to	Barcelona
		‘Travel to Barcelona!’
	c.	* Er	fährt	denn	nach	Barcelona.
		he	travels	part	to	Barcelona
		‘He is traveling to Barcelona.’

(35) demonstrates that modal particles are restricted to specific sentence types. In our case, denn can only occur in interrogatives (35a) and not in imperatives (35b) or declaratives (35c). However, even within a particular sentence type, modal particles are confined to specific use conditions. Look at a prominent example by König (1977, 119):

(36)	CONTEXT: A wakes up B and A asks:
	# Wie	spät	ist	es	denn?
	how	late	is	it	part
	‘What time is it?’

According to the literature, one component of denn is to indicate that the information that is asked for by the speaker is part of a discourse that has already been established between the speaker and the addressee. The example in (36) illustrates that questions featuring denn are infelicitous when the addressee (here: B, who just woke up) lacks a context (read: common ground) in which to interpret the question. However, recently it has also been noted in the literature (e.g., Theiler 2018) that denn-questions can actually be asked out of the blue:

(37)	CONTEXT: Someone asking a passerby:
	Wo	ist	denn	hier	der	Bahnhof?
	where	is	part	here	the	train-station
	‘Where is the train station here?’

It has thus been suggested that the obligatory relation to the discourse that denn expresses must thus be conceived in broader terms, and that denn rather is an abstract ‘discourse-navigating device’ (Zobel and Csipak 2017) that ‘helps interlocutors with navigating a discourse’ (Theiler 2018). In recent work on the Dutch cognate of German denn, the particle dan, Trotzke and Haegeman (in press) suggest picturing the abstract discourse function of the German particle denn as follows:

(38)	Wo	ist	denn	der	Bahnhof?
	where	is	part	the	train-station
	‘Where is the train station?’

The main point by Trotzke and Haegeman (in press) is that given their data on both Dutch dan and German denn, they take the informal paraphrase of denn as a ‘discourse navigating device’ seriously by endorsing a more narrowly spatial interpretation of denn. That is, they conceive denn as a navigation device between figure and ground, and an utterance like (38) would encode the meaning depicted in Fig. 1, where denn navigates between the ‘ground’ (the reason(s) for posing the question) and the ‘figure’ (the question/the highlighted, focused part of the question). Trotzke and Haegeman (in press) argue that such a core meaning of denn is supported by its grammaticalization path that has been postulated in the diachronic literature (see Abraham 1991; Wegener 2002):

Fig. 1

denn navigating between the ‘ground’ (the reason(s) for posing the question) and the ‘figure’ (the question/the highlighted, focused part of the question)

(39)	localistic > temporal > logical > illocutive/discourse functional
	(Abraham 1991, 373)

In particular, the particle denn can be derived from Idg. root *to, which expresses a deictic meaning. OHG danne/thanne (cf. English then) first was a locative adverb and only later became a temporal adverb.

Be that as it may, the data above show that modal particles are not only restricted to different sentence types and their general illocutionary potential. Rather, the use of modal particles is also confined to certain illocutionary subtypes expressed by the same sentence type and can teach us something about the use conditions of these subtypes.

In other words, the German versions of A-EXCs are interesting because the occurrence of modal particles such as denn can help identify the discourse meaning and exact use conditions of these utterance types. Based on the sketch above, we can say that denn signals reference to the common ground between speaker and addressee. Now, German A-EXCs can also contain other particles that share some crucial features with denn. Here are some examples featuring the particle eigentlich (lit. ‘actually’):

(40)	Aber	zurück	zum	Clip:	Wie	geil	ist	der	eigentlich?	Ich	war
	but	back	to.the	clip	how	cool	is	it	part	I	was
	stolz	wie	Oskar,	als	er	veröffentlicht			wurde.
	pride	like	Oscar	when	it	released			was
	‘But back on the clip: How cool was that? I was as proud as a peacock when it got released.’
	<attested at http://www.pleasuremag.com on March 1, 2019>

(41)	Meine	Familie	und	Freunde		waren	entsetzt	und	jeder	hat	sich
	my	family	and	Friends		were	shocked	and	everyone	has	themselves
	gefragt	„Wie	blöd	ist	der	eigentlich?!“.		Denn	ich	stand	nun
	asked	how	stupid	is	he	part		because	I	stood	now
	da,	ohne	Abschluss		mit	abgebrochener		Berufsausbildung.
	there	without	degree		with	incomplete		vocational.training
	‘My family and friends were shocked, and everyone asked themselves: How stupid is he! Because now I stood there without a degree and with incomplete vocational training.’
	<attested at https://www.muskelkater-giessen.de on March 1, 2019 >

Note that eigentlich is in principle ambiguous between a modal-particle and an adverb reading, which is typical for many German particles. One way to clearly distinguish the two readings is placing stress on eigentlich, since modal particles in questions cannot be stressed:

(42)	a.	Wie	blöd	ist	der	eigentlich!
		how	stupid	is	he	part
		‘How stupid is he!’
	b.	Wie	blöd	ist	der	EIgentlich?
		how	stupid	is	he	actually
		‘How stupid is he actually?’

In the examples (40) and (41) above, eigentlich clearly occurs in its particle-reading and thus could not receive heavy stress. Another particle occurring in German versions of A-EXCs is bitte (lit. ‘please’)⁹:

(43)	Omg!	Ich	brauch	dieses	Hausschwein	von	Marcel	Wanders!	Wie	geil
	OMG	I	want	this	squealer	by	Marcel	Wanders	how	cool
	ist	das	bitte?!
	is	that	part
	‘OMG! I really want this squealer by Marcel Wanders! How cool is it!’
	<attested at http://www.imgrum.net on March 1, 2019>

(44)

[…]

und

freitag

fragt

er

K

ob

wir

in

fb

befreundet

sind?!

hallo?!

and

Friday

asks

he

K

whether

we

in

FB

friends

are

hello

wie

blöd

ist

das

bitte?!

how

stupid

is

that

part

‘[…] and Friday he asks K whether we are friends on Facebook? Come on! How stupid is that!’

<attested at http://www.maedchen.de on March 1, 2019>

Crucially now, we observe that German A-EXCs can only contain a subset of particles that can occur in German wh-questions (see Thurmair 2013 and Table 1 above). For instance, consider the cases of nur (lit. ‘only’) and schon (lit. ‘already’).

Table 1 Distribution of modal particles in wh-exclamatives and wh-questions

The particle nur does not refer to the common ground between speaker and addressee in the way denn, eigentlich, and bitte do. In (45a), nur signals that the speaker has tried a number of times to find an answer, but without success (see Dörre and Trotzke 2019 for semantic details). Similarly, the particle schon also differs from the particles introduced above because it serves to mark rhetorical questions (45b); see Bayer and Obenauer (2011); Trotzke and Turco (2015):

(45)	a.	Wie	geil	ist	das	nur?
		how	cool	is	that	part
		‘How cool is that? (I have already asked myself several times.)’
	b.	Wie	geil	ist	das	schon?
		how	cool	is	that	part
		‘After all, how cool is that? (Not very cool!)’

Both particles are ruled out in the A-EXC interpretation, as (46) illustrates. Finkbeiner (2015, 247–248) provides empirical evidence from Google searches that confirms this impression: the particles wohl, nur, bloß are virtually excluded from German How cool is that.

(46)	CONTEXT: A is completely amazed at the musician’s performance and would like to share his amazement with B who is attending the show together with A. A exclaims:
	a.	??	Wie	geil	ist	das	nur!
			how	cool	is	that	part
‘How cool is that? (I have already asked myself several times.)’
	b.	??	Wie	geil	ist	das	schon!10
			how	cool	is	that	part
			‘After all, how cool is that?’ (Not very cool!)’

On the other hand, denn and eigentlich do not belong to the class of exclamative particles. These particles, which frequently occur in A-EXCs, are restricted to interrogative syntax and are thus incompatible with wh-exclamatives:

(47)	*Was	wir	denn/eigentlich	für	blöde	Kerle	sind!
	what	we	part/part	for	stupid	guys	are
	(intended: ‘What stupid guys we are!’)

This is corroborated by Thurmair’s (2013, 638) comprehensive overview, which mentions that the following particles (in bold) are licensed in wh-exclamatives (this list is consistent with the inventory mentioned by d’Avis 2016, 162):

(48)	Was	wir	doch/aber auch/bloß/nur	für	blöde	Kerle	sind!
	what	we	part/part/part/part	for	stupid	guys	are
	‘What stupid guys we are!’

As has also been observed for the ‘minor’ sentence type of optative constructions (see Grosz 2012), the set of modal particles is relatively small in exclamatives, and the literature has not proposed any significant meaning differences depending on what kind of particle is used (however, see some remarks on subtle differences in Thurmair 1989, 190–194). One could speculate that the set is rather small because wh-exclamatives have enough distinct properties and thus do not depend on modal particles as cues as much as other sentence types do. As already mentioned above, specific lexical markers (such as what a or was für in German), together with exclamative intonation, are clear indications of an exclamatory interpretation.

Table 2 Distribution of modal particles in German A-EXCs (in bold)

Let us summarize our observations by looking at Table 2, a revised version of Table 1. The table indicates that the set of possible particles in German A-EXCs does not correspond to an intersection between particles that are licensed in wh-questions and particles that occur in exclamatives. Rather, the special pragmatic function of A-EXCs is signaled by a dedicated class of modal particles, namely denn, eigentlich, and bitte. The question thus arises whether the question particles schon, wohl, and nur/bloß share a common property that excludes them from the A-EXC reading on the one hand, and if, on the other hand, denn, eigentlich, and bitte share a property that explains why they are felicitous in the A-EXC reading.

Let us first turn to the question of what common property the particles schon, wohl, and nur/bloß share that excludes them from A-EXCs. I already mentioned above that schon can be used to signal that a wh-interrogative should be interpreted as a rhetorical question. Note that particles like schon are not obligatory to form ‘special’ (here: rhetorical) questions; but once these particles are used, they serve as reliable ‘communicative cues’ (Grosz 2014) to indicate to the addressee that the interrogative should be interpreted as a rhetorical question:

(49)	Wie	hoch	ist	das	schon?
	how	tall	is	that	part
	‘After all, how tall is that?’ (Not very tall!)’

One of the main characteristics of a rhetorical question is that an answer to that question is not expected. Now, how do wohl and nur/bloß connect to this type of interpretation that licenses the use of schon? Consider again an example featuring these particles:

(50)	Wie	hoch	ist	das	wohl/nur/bloß?
	how	tall	is	that	part/part/part
	‘How tall is that?’

Note that it is striking that exactly these three particles (and only these) are obligatory in wh-verb-final sentences in German that express a deliberative and/or speculative interpretation (see Truckenbrodt 2013, 241):

(51)	a.	Wen	sie	wohl/nur/bloß	mag?
		who	she	part/part/part	likes
		‘Who does she like?’
	b.	# Wen sie mag?
		[only possible reading: echo question]

These deliberative questions are typically not addressee- but rather self-oriented. Accordingly, speakers do not expect to receive a straight answer and are mulling over something that is currently uppermost in their mind. What these questions and the use of the particles wohl, nur, and bloß thus share with schon in rhetorical questions is that they express:

‘that the obligation to answer the question is suspended.’
(common discourse feature of German particles wohl, nur, bloß, and schon)

Let us now look at the particles that are felicitous (if not virtually obligatory) in How cool is that (i.e., denn, eigentlich, and bitte). I already referred to the basic meaning of denn above by citing the prominent example from König (1977), repeated here for convenience:

(52)	A wakes up B and A asks:
	# Wie	spät	ist	es	denn?
	how	late	is	it	part
	‘What time is it?’

As already highlighted above, denn indicates that the reason for posing the question can be found in the current discourse context (visualized in Fig. 1 above). (53) is infelicitous because the addressee (B, who just woke up) lacks a discourse context in which to interpret the question. Note now that eigentlich is infelicitous for the same reason; it requires that the addressee is aware/has knowledge about the current discourse context:

(53)	A wakes up B and A asks:
	# Wie	spät	ist	es	eigentlich?
	how	late	is	it	part
	‘What time is it?’

This observation is in accordance with work by Eckardt (2009) who highlights that unstressed eigentlich (i.e., the particle use of eigentlich) is infelicitous in questions in certain discourse-initial environments (see Eckardt 2009, 104–105). Concurring observations have also been made for Dutch eigenlijk (van Bergen et al. 2011), namely that this particle, in van Bergen et al.’s terms, expresses that the speaker is aware of assumptions shared with the addressee.

Crucially now, the common semantics of denn and eigentlich cannot be the whole story of why these particles are felicitous in German A-EXCs. Turning to the particle bitte, we observe that this particle is perfectly fine in the context above:

(54)	A wakes up B and A asks:
	Wie	spät	ist	es	bitte?
	how	late	is	it	part
	‘What time is it?’

Zimmermann (2009) points out that bitte is only infelicitous when the context is such that neither speaker nor addressee are in a position to answer the question (and even B, who just woke up, is at least able to answer the question, if not necessarily prepared to answer at that very moment). To see this, look at the following examples taken from Zimmermann (2009, 57); note that the point of verb-final ob-interrogatives (55b) is to express mutual ignorance:

(55)	A and B are lost in the woods while hiking. A asks B:
	a.	# Welches	ist	bitte	der	Weg	aus	dem	Wald?
		which	is	part	the	way	out	the	forest
		‘What would be the way out of the woods?’
	b.	# Ob	das	bitte	der	richtige	Weg	ist?
		if	this	part	the	right	way	is
		‘Will/Would this be the right way?’

The particles denn and eigentlich are ruled out in the same context:

(56)	a.	# Welches	ist	denn/eigentlich	der	Weg	aus	dem	Wald?
		which	is	part/part	the	way	out	the	forest
		‘What would be the way out of the woods?’
	b.	# Ob	das	denn/eigentlich	der	richtige	Weg	ist?
		if	this	part/part	the	right	way	is
		‘Will/Would this be the right way?’

We can thus conclude that the common property of denn, eigentlich, and bitte is that they are only licensed in a context where

‘A believes that B is able to supply an answer to the question.’
(common discourse feature of German particles denn, eigentlich, and bitte)

This basic component of the particles’ meanings has already been formulated for denn by Zobel and Csipak (2017), but the examples above illustrate that it also holds for eigentlich and bitte. In other words, only those particles that express that the addressee is able to supply the requested response, given an already established common ground between speaker and addressee, are felicitous in A-EXCs. Note that the use of schon, wohl, and nur/bloß are not subject to the same felicity condition: in the case of wohl and nur/bloß, for instance, the speaker can only hope that B is able to supply an answer, but the point of using nur/bloß is that A cannot be sure about B’s ability to provide an answer (in fact, both particles leave open the possibility that nobody, including A and B, can answer the question).

Interestingly, languages that do not feature modal particles often express the meaning contribution of particles like denn, eigentlich, or bitte by other linguistic means/phrases. These alternative lexical or phrasal means also convey that the speaker and addressee are already part of a shared discourse, and that their common ground allows for the conclusion on the part of the speaker that the addressee is perfectly able to supply an answer to the question. Consider the following selection of English examples gathered via Google search:

(57)	Heck, they can play with the software themselves to add whatever they want to the lifestyle search. I mean, come on, how cool is that? […]”
	<attested at http://blog.spatialmatch.net on March 1, 2019>

(58)	There is no project that upstream wants to implement this hack, and finally what is cooking, a brand new font! Now, how cool is that!
	<http://gregdekspeaks.wordpress.com/2010/07/29/red-hat-16-canonical-1>

(59)	At INVERS, I developed an app with which I can open my car… I mean, how cool is that?!
	<https://invers.com/company>

(60)	Seriously, how cool is that? Using a camera and a projector to learn how to play bass guitar! Augmented Reality here we come
	<https://www.reddit.com/r/technology/comments/94yr6/seriously_how_cool_is_that_using_a_camera_and_a/>

(61)	Well Now, How Cool is That? Today was my first time to enter the Camera Critters meme.
	<http://snowbirdblogspotcom.blogspot.com/2008/09/well-now-how-cool-is-that.html>

Of course, these examples and observations are only preliminary and therefore deserve further in-depth investigation. Importantly, these preceding lexical/phrasal means to indicate that the speaker has reasons for assuming that the addressee shares assumptions about the ongoing discourse and is thus able to provide a response cannot felicitously occur with the exclamative (= verb-final) counterparts of A-EXCs:

(62)	a.	??	Seriously, how cool that is!
	b.	??	I mean, come on, how cool that is!
	c.	??	Well now, how cool that is!

Note that discourse-marking elements like in (57)–(61) are also often used in German in addition to modal particles. For example, example (44) above additionally contained a special use of German Hallo, preceding the utterance. It is pronounced with a rising contour [/] and also exists in English when used to express indignation and/or annoyance; compare the following German example:

(63)	Ich	bin	hier	wohl	im	falschen	Film!	Hallo!	Geht’s	noch?
	I	am	here	part	in.the	wrong	movie	hello	goes = it	still
	‘I guess I’m in the wrong movie here! Hello? Are you crazy?’
	<https://www.bod.de/autorenpool/meinung-gefragt-t16661,start,15.html>

It would be odd if this element would be used as preceding an exclamative, since it conveys a certain subjective shakiness on the part of the speaker, which is in conflict with strong and thus confident evaluations expressed by exclamatives. In a sense, elements such as Hallo!?, seriously?!, etc. are anchored with the addressee and elicit a certain kind of answer (a serious one, a straight one, etc.). It is thus no surprise that this ‘interrogative flip’ is incompatible with exclamatives (on this notion, see Tenny 2006).

Let us take stock. Although A-EXCs and wh-exclamatives in both English and German share their exclamation component (see Sect. 2 above), looking at the German data suggest that A-EXCs ask for affirmation by the addressee to a greater extent. In particular, this is indicated by response-evoking modal particles in German (and analogous discourse marking elements in English, as the data set above suggests). My terminological distinction, already introduced at the outset of this paper, is thus corroborated by the data discussed above: A-EXCs aim at a reaction by the addressee, whereas exclamatives, on the other hand, do not aim at an affirmation (or, for that matter, at an objection) on the part of the addressee. Rather, the expressive content (‘that the speaker is surprised about p’) is a subjective emotion on the part of the speaker, and the addressee can of course affirm or object to this emotion, but the utterance of an exclamative does not require such a reaction or response by its linguistic form. We thus find two subtypes in the domain of exclamation speech acts (EXC): one that asks for affirmation on the part of the addressee (A-EXC) and one that does not, and which I would like to call ‘speaker-oriented’ exclamations (or S-EXC for short).

This is in line with what has recently been proposed for another class of inversion exclamatives in English. In particular, while Taniguchi (2017, 127–135) claims that all exclamatives are only about speaker commitment (and thus ‘not up for discussion’), Jamieson (2018) has shown that at least negative inversion exclamatives must also involve some connection with (the evidential situation of) the addressee. Consider the following example from Jamieson (2018, 179):

(64)	S is on the phone to A, and is telling A about their new cat. S just picked up the cat that day. S says:
	# Isn’t it cute!

(64) cannot be explained by a model like Taniguchi’s (2017) where the speaker—when uttering any type of exclamative—updates the Common Ground without inquiring what the addressee thinks. We thus see that there are even more syntactic configurations that support the idea that we must distinguish between A-EXCs and S-EXCs, as introduced above. I must admit that I do not have an explanation for why cases like How cool is that and negative inversion exclamatives behave similarly in this regard. Note that it cannot be due to the inversion property because positive inversion exclamatives (i.e., Boy, is it cute!) would be totally fine in a context like (64). My hunch is that for the exclamatives with polar question syntax, the answer lies in the non-degree interpretation of negative inversion exclamatives (as discussed in length by both Taniguchi 2017 and Jamieson 2018). However, as for the wh-cases of A-EXCs, I will argue that their addressee-orientation can indeed be derived from their word order.¹¹ Accordingly, in order to capture the full pragmatics of A-EXCs, we have to complement the idiosyncratic properties sketched in Sect. 3.1 and often highlighted in the constructionist literature with relevant aspects of compositional syntax.

3.3 The pragmatics of A-EXCs and the V-to-C property

Finkbeiner (2015) points out that, in contrast to regular wh-questions, A-EXCs are not conveying the expectation that the addressee should provide an answer. Crucially, however, she also hypothesizes something along the lines mentioned in the last section above: that A-EXCs come with an obligation to react to them.¹² Related to this characterization, she postulates that A-EXCs are essentially evaluative acts: that is, they can express different evaluations, depending on what kind of adjective is used; the following examples are given by Finkbeiner (2015, 267; translations are mine) and are based on her data gathered from Google:

(65)	a.	Wie	smart	ist	das	denn!	[admiration]
		how	smart	is	that	part
		‘How smart is that!’
	b.	Wie	krank	ist	das	denn!	[contempt]
		how	sick	is	that	part
		‘How sick is that!’
	c.	Wie	lecker	ist	das	denn!	[appreciation]
		how	yummy	is	that	part
		‘How yummy is that!’

Finkbeiner (2015) points out that evaluative acts in general are characterized by the addressee-oriented goal to prompt the addressee to share the evaluation and/or to recommend or advise against something. I agree with Finkbeiner (2015) that this addressee-oriented pragmatics holds for cases like (65). It is not clear to me, however, if and how A-EXCs (65b) then differ from other wh- or also nominal exclamations (66a) because such a difference is what Finkbeiner (2015) claims in the first place:

(66)	a.	What a tall building!/Was für ein hohes Gebäude!
	b.	How tall is that building!/Wie hoch ist das Gebäude denn!

It is not obvious that the exclamations in (66a) aim at a reaction by the addressee although they should also be classified as ‘evaluative acts’, according to Finkbeiner’s (2015) discussion. In what follows, I will show that in fact only A-EXCs of the form given in (66b) clearly display such a discourse function.

Importantly, this is supported by both the distribution of modal particles discussed in the last section and the property of V-to-C movement of such cases. In particular, while the exclamation component of such questions (discussed in Sect. 2) is not triggered by word order, its special interpretation of asking for a response from the addressee is. To show this, I will adopt the theory of V-to-C movement by Truckenbrodt (2006), certainly one of the most detailed approaches to the semantic and pragmatic effects of this syntactic operation.

Let us start by looking at the following German main clauses. In (67a), involving stress-bearing JA, the main clause does not exhibit V-to-C movement, while (67b) features this syntactic operation (see Truckenbrodt 2006, 269):

(67)	a.	Dass	du	(JA)	das	Fenster	öffnest!
		that	you	(part)	the	window	open
		‘(Don’t forget to) open the window!’
	b.	Du	öffnest	das	Fenster.
		you	open	the	window
		‘You’re opening the window.’

Although the verb is inflected for second person singular in both (67a) and (67b) and although both examples denote a proposition, (67a), lacking V-to-C movement, cannot express an assertion, but is used here as a directive speech act. According to Truckenbrodt’s (2006) theory, the difference between these two word orders can be captured in terms of an epistemic component residing in the C-domain of the clause (in what follows: Epist). To see this, consider the following speech act paraphrases of a declarative (68a) and a corresponding interrogative (68b), both featuring V-to-C movement (see Truckenbrodt 2006, 264–265); S = speaker; A = addressee:

(68)	a.	Der	Peter	hat	das	gemacht.
		the	Peter	has	this	done
		‘Peter has done this.’
		Speech act paraphrase: ‘S wants from A that it is common ground that Peter has done this.’
	b.	Hat	der	Peter	das	gemacht?
		has	the	Peter	this	done
		‘Has Peter done this?’
		Speech act paraphrase: ‘S wants from A that it is common ground whether Peter has done this.’

Note that Truckenbrodt postulates that the interpretations of both declaratives (68a) and interrogatives (68b) always have the common ground as the ‘epistemic desideratum’, in his terms. Accordingly, in both cases of V-to-C movement above, the feature Epist yields the meaning component ‘it is common ground that/whether…’ Note also that Truckenbrodt (2006) conceptualizes “all sentential speech acts as volitional on the part of the speaker […] S wants something, wishes for something, invites A to do something etc.” (Truckenbrodt 2006, 263–264). Given this deontic conception of the speaker’s intentions (abbreviated as ‘Deont_S’), Truckenbrodt argues that the C-head contains a set of features that can be formally captured as a ‘context index’ (Truckenbrodt 2006, 265):

(69)	In a context index 〈DeontS (,x) (,〈Epist〉)〉 in C
	a. Epist is present iff
	(i) C contains a finite verb with indicative or Konjunktiv II
	or
	(ii) C/CP is marked [+ WH];
	b. x = A iff C contains a finite verb with person inflection.

We can easily see why Truckenbrodt includes (ii) in (69a): root interrogatives without V-to-C movement but with [+ WH] can feature the epistemic interpretation (Truckenbrodt 2006, 274), in contrast to cases like (67a) above.

(70)	[+ WH], no V-to-C movement
	Ob	er	immer	noch	kubanische	Zigarren	mag?
	whether	he	always	still	Cuban	cigars	likes
	‘I wonder whether he still likes Cuban cigars?’
	‘S wants it to be common ground whether he still likes Cuban cigars.’

Although [+ WH] configurations can thus have Epist without V-to-C movement, these structures lack the addressee-oriented component ‘from A’, which is present in the interpretations of (68a) and (68b) above. In other words, questions like (70) are typically used when the speaker assumes mutual ignorance, that is, when he does not expect the addressee to know the answer; see also our discussion of Zimmermann’s (2009) example in Sect. 3.2 above. Hence, according to Truckenbrodt, [+ WH] alone cannot provide the value ‘A’ for ‘x’. Only if V-to-C is present, the value A can be provided in the context index of the C-domain.

Coming back to our exclamatory cases, I agree with Truckenbrodt (2006) that exclamatives are outside this deontic-epistemic system. As I already indicated in Sect. 2, various sentence types can receive an exclamatory interpretation by intonational, lexical means, and emphatic fronting of various phrasal constituents to the left periphery. And, at first sight, also our cases of A-EXCs are a good example of this for the most part unrestricted situation, given that the heavy stress on the subject is prototypically associated with the exclamation interpretation:

(71)	a.	How tall is THAT!
	b.	How TALL is that?

However, while wh-exclamatives have neither Deont_S, Epist, nor A, and thus do not necessarily feature V-to-C movement (as is the case in English and German), utterances of the form How cool is that! or Wie geil ist das denn! cannot be characterized as wh-exclamatives, although they share the exclamation component with their exclamative counterparts. Once we acknowledge that A-EXCs slightly differ in their pragmatics from wh-exclamatives (indicated in German by a dedicated set of particles; see Sect. 3.2 above), it makes sense in the light of Truckenbrodt’s observations that they necessarily involve V-to-C movement.

At this point, we have to recall from Sect. 3.2 above that wh-exclamatives in German can also feature T-to-C movement. Consider, again, the following examples:

(72)	Wie	vergesslich	ist	Oma	aber (auch)	seit	letztem	Jahr	geworden	ist!
	how	forgetful	is	grandma	part	since	last	year	gotten	is
	‘How forgetful grandma has gotten since last year!’

(73)	Wie	vergesslich	ist	Oma	denn/wohl	seit	letztem	Jahr	geworden	ist?
	how	forgetful	is	grandma	part/part	since	last	year	gotten	is
	‘How forgetful has grandma gotten since last year?’

German thus features structurally ambiguous wh-configurations with T-to-C movement. However, the structures in (72) and (73) can be disambiguated by specific particles: in (72), the exclamative particle aber (auch) clearly signals that the structure should be interpreted as an exclamative, and thus the particle serves as a ‘communicative cue’ (in the sense of Grosz 2014) for the exclamation reading.

On the other hand, the question particles in (73), denn and wohl, signal that this T-to-C-movement configuration should be interpreted as a regular (i.e., information-seeking) wh-question. Given this situation, we can say that the interrogative reading of German wh-structures involving T-to-C-movement is always a combination of different linguistic means: T-to-C-movement, the use of certain question particles that can only be used in T-to-C-movement structures (like denn or wohl above), and non-exclamative question prosody.

In a way, T-to-C-movement and question particles add up to the A-EXC reading in German. This is why empirical work on German A-EXCs has shown that the question particles are virtually obligatory in the German version of How cool is that! (see Auer 2016 and our discussion in Sect. 3.2 above). In English, on the other hand, verb movement and subject inversion does the job of yielding the interrogative (the addressee-oriented) component of A-EXCs like How cool is that!, which the corresponding configurations without T-to-C movement (How cool that is!) lack. In other words, since English lacks T-to-C-movement in wh-exclamatives, there is no need in this language for extra cues signaling that How cool is that! is an A-EXC and not an exclamative (and thus a S-EXC).¹³

After having pointed out this important qualification, let us now return to the central claim in (69) above. According to (69), we can assume that the pragmatic difference in terms of illocutionary force between wh-exclamatives lacking V-to-C movement and A-EXCs featuring V-in-C is captured by the fact that only A-EXCs have a context index that contains the addressee-related value ‘A’ for ‘x’. In German, as we have pointed out above, further disambiguating devices may be used in order to signal that V-in-C structures are indeed interpreted as interrogatives and thus do not pattern with verb-final exclamatives in lacking Deont_S, Epist, and A altogether.

In sum, then, we can conclude that when speakers utter A-EXCs, the corresponding speech act can be paraphrased as in (74); on the degree component, see the semantics of [wh + Adj] in Sect. 2.2:

(74)	‘S wants from A that it is common ground whether the degree to which p holds goes beyond a threshold θ.’

In other words, A-EXCs have a deontic-epistemic interpretation that wh-exclamatives lack and that they share with interrogatives. The derivation, according to this analysis and what we have claimed for the exclamation component in Sect. 2, can thus be represented as follows:

(75) spells out the claim that A-EXCs are in fact interrogatives with an additional exclamation component that can be analyzed along the lines of other cases where phrasal constituents receive a scalar reading by fronting them to the specifier of EmpP before they must move further for reasons of clausal typing. Note again that EmpP constitutes a syntactic layer that cannot be identified with the function of clausal typing and/or illocutionary force. More specifically, Trotzke (2017a) has argued that emphasis in questions is a type of not-at-issue meaning whose operator must be below the Question operator for many empirical reasons (see Sect. 2 above).

The analysis for German A-EXCs (e.g., Wie geil ist das denn!) would look the same, except for the placing of the modal particle denn in the highest position of the middle field (IP zone); see Bayer and Obenauer (2011) and Bayer and Trotzke (2015) for extensive discussion and detailed derivations of modal particles in questions. The licensing of only a subset of question particles in A-EXCs can be explained by referring to both the illocutionary point of A-EXCs and the specific use conditions of these particles formulated in Sect. 3.2 above. In other words, the illocutionary meaning of A-EXCs in (76) is only compatible with particles that contribute the use condition in (77a), but it is incompatible with the use condition given in (77b):

(76)	‘S wants from A that it is common ground whether the degree to which p holds goes beyond a threshold θ.’

(77)	a.	use condition for denn, eigentlich, bitte:
		‘A believes that B is able to supply an answer to the question.’
	b.	use condition for wohl, nur, bloß, schon:
		‘that the obligation to answer the question is suspended.’

I follow Gutzmann (2015) and many others here in claiming that those incompatibilities should not be modeled in terms of syntactic mechanisms, but can be explained on purely pragmatic grounds (see also Grosz to appear for a general overview of this line of thought). Given this analysis of A-EXCs, let us now summarize the paper and also turn to the question again of how we should deal with all the remaining idiosyncratic properties of A-EXCs that we discussed in this paper.

4 Conclusion

In this paper, I focused on a subclass of non-standard uses of interrogatives, namely on special uses of how-interrogatives. After having discussed some general issues concerning the connection between sentence type and illocutionary form, I presented the new observation that utterances like English How cool is that! have in fact a hybrid status: they function as both speaker-oriented exclamations and addressee-oriented questions. Drawing on cross-Germanic evidence from English and German, I specified their pragmatic contribution as addressee-oriented exclamation and thus labeled these wh-configurations with interrogative form as A-EXCs. A-EXCs can be signaled by a dedicated class of modal particles in German, and their meaning difference to wh-exclamatives is clearly expressed by word order in both German and English.

Taken together, the evidence discussed in this paper shows that these wh-configurations have a unique pragmatic function, and that this function is indicated by a combination of non-syntactic means and compositional syntax. As for syntactic work, I agree that A-EXCs are of course characterized by a complex interplay of intonational, syntactic, and lexical factors. A construction-oriented analysis that can assemble these diverse aspects under one ‘construction’ thus seems to be a natural account at first sight.

And indeed, a construction-based view of A-EXCs could, in principle, also account for the V-to-C-movement component of A-EXCs. Consider the following construction-based analysis of German V2-wh-questions by Jacobs (2016, 37), where ‘FinP’ is the Finiteness Phrase (and the propositional core of the clause), ‘-sub’ stands for ‘-subordinated’, ‘prop’ is the proposition, and ‘ill-typ’ the illocutionary type:

According to a construction-based perspective, one could then argue that the characteristics of a general ‘V-to-C-construction’ (or here: a ‘V-to-C-construction’ involving a wh-element in SpecCP) could just be inherited by more specific constructions such as the ‘A-EXC construction’. However, note that this would not be a case of ‘strict inheritance’ in the sense of Flickinger et al. (1985), where all properties of the general construction are inherited by the more specific construction.

More specifically, our discussion in the paper has shown that wh-questions and A-EXCs are still distinguished by, for instance, some intonational choices (e.g., heavy stress on the subject d-pronoun can in fact disambiguate between wh-question and A-EXC; see Sect. 3.1 above). Although one could then model these weaker relations between wh-question and A-EXC by using other formal tools (see, e.g., Sag’s 2010 approach to English sentence types within Sign Based Construction Grammar and Boas and Ziem 2018 for a general constructionist approach to German syntax), my claim in this paper has merely been that, whatever formal model one would use, the connection between syntactic form and meaning in A-EXCs is by no means arbitrary and unrestricted. Since arbitrary form-function mappings are a cornerstone of a constructionist view on language, I argue that a compositional conception of sentence-type meanings can better explain what we observe in the case of minor sentence types such as How cool is that!

All in all, instead of giving up the idea of formulating a systematic relation between syntactic form and illocutionary potential (that also includes ‘minor’ types such as A-EXCs), I argued in this paper that such a compositional understanding of sentence meaning is relevant even for pragmatic uses that at first seem rather marginal and formally unrestricted. I hope to have shown that both systematic structural features and non-formal (sometimes also idiosyncratic) components are relevant for our understanding of syntactic constructions—also for the new ones that keep emerging continuously.