Prospective Aspect and Current Relevance: A Case Study of the German Prospective Stehen vor NP Light Verb Construction

Abstract

The paper presents a detailed corpus-based analysis of the German prospective stehen vor NP light verb construction. The starting point of the analysis is the claim that the construction is restricted to change-of-state nouns in the NP-internal position (Fleischhauer & Gamerschlag 2019, Fleischhauer et al. 2019). Based on corpus data, I demonstrate that although the construction shows a strong preference for such nouns, other semantic types of nouns (such as state nouns or process nouns) occur in the construction as well. I argue that process nouns in particular require contextual support to be licensed within the construction. In the paper, I present an analysis of the prospective light verb construction in terms of current relevance. This analysis accounts for the observed preference for change-of-state NP-internal nouns as well as for the need to provide contextual support for process nouns. The notion current relevance is frequently employed in the analysis of the perfect aspect; the current paper represents the first attempt to extend this notion to the prospective aspect.*

1. Introduction.

Grammatical, or viewpoint aspect is a grammatical category “concerned with different ways of representing the internal temporal constitution of a situation” (Comrie 1976:3). In a narrow sense, the category aspect consists of the two values perfective and imperfective and their subtypes, such as progressive or habitual. Most Germanic languages do not have obligatory expression of aspect within their verbal paradigms and therefore lack grammatical means for it. An exception among the Germanic languages is English, which has a grammaticalized expression of progressive aspect.

Whether the perfect and the prospective should also be included under the label aspect is disputed in the literature (see the discussion in Ritz 2012). It is often stated that the perfect (as well as the prospective) combine temporal and aspectual features (see, for instance, the discussion in de Swart 2012:762, Grønn & von Stechow 2020 and references cited therein). However, even those authors who subsume the perfect and the prospective under the label aspect admit that these types of aspect differ from the one encoding the perfective/imperfective opposition. Dik (1997:221) refers to the latter as “aspect proper” and speaks of the perfect and the prospective as instances of “perspectival aspect.” The meaning of neither the perfect nor the prospective can exhaustively be characterized in purely temporal terms (as I show below for the prospective), and therefore I refer to the two categories as aspectual in the wider sense of Dik 1997.¹

Although there exists an extensive literature dealing with the perfect, the prospective aspect has received considerably less attention. With respect to German, the prospective aspect is occasionally mentioned (Harweg 1976, Thieroff 1992, among others) but rarely discussed in detail. Two exceptions are Fleischhauer & Gamerschlag 2019 and Fleischhauer et al. 2019, which analyze the prospective stehen vor NP light verb construction. An example of this construction is shown in 1. The subject referent die Firma ‘the company’ will, if nothing intervenes, undergo a change, namely, it will be handed over to a manager.

1. (1)

   

In their analysis of this construction, Fleischhauer & Gamerschlag as well as Fleischhauer et al. propose that in 1, the noun within the NP—Übergabe ‘handover’—denotes a change of state. So far, this restriction has not been subject of any empirical investigation. In the current paper, I take up this issue and present the results of a corpus study on the prospective stehen vor NP construction. The corpus study has been carried out with the intention of identifying the range of nouns licensed within the construction. As I demonstrate in the paper, the prospective construction does not impose any lexical restrictions on the semantic type of the nouns realized within the vor PP. However, the construction is not entirely unconstrained. Examples such as 2 are judged as unacceptable by native speakers if presented without a proper context.

1. (2)

   

I argue that the restrictions are of pragmatic rather than semantic nature. Adopting a suggestion by Fleischman (1982, 1983), I develop an analysis of these restrictions in terms of current relevance. I argue that a noun is only licensed in the prospective light verb construction if the eventuality denoted by the NP can be interpreted as being of current relevance. This analysis is supported by the fact—as I demonstrate in section 6—that some nouns (such as Spiel ‘game’) require contextual support to be licensed within the construction. I propose that based on context, an eventuality can acquire current relevance by virtue of being, for instance, important or challenging. Concerning the example in 2, I argue that a future event of bleeding does not have current relevance without a wider context: Bleeding is a process noun that on its own does not entail any particular result; neither is the event of bleeding particularly challenging for the subject referent. The notions important and challenging are further discussed in section 6, where this issue is addressed.

Current relevance is a central notion in the discussion of the (resultative) perfect, which portrays the result of a past event as being relevant for the current moment. To the best of my knowledge, so far, the notion current relevance has not been employed to account for properties of the prospective aspect. The present paper aims at filling this gap. I further explore this notion using the corpus data presented in the paper.

The paper is structured as follows: Section 2 starts with a general discussion of the prospective aspect before illustrating its different periphrastic realization patterns in German. In section 3, I present a corpus study of the construction illustrated in 1. The aim of the corpus study consists in identifying the range of nouns admissible within the construction’s NP slot. A semantic classification of the nouns found within the construction is provided in section 4. In section 5, I explore the question of whether different instances of the analyzed prospective construction have a unified interpretation. I argue that they do, if one separates the lexical meaning of the embedded nominal predicate and the meaning of the prospective construction as a whole. Finally, in section 6, I turn to a discussion of the notion current relevance.

2. Prospective Aspect.

The prospective aspect, sometimes also referred to as “proximative” (Heine 1994, Kuteva 2001, Kuteva et al. 2019), expresses a relation between the subject argument’s current state and a subsequent eventuality (Comrie 1976:64). An illustrative example is The ship is about to sail (Comrie 1976:64), which presents the ship as being in a state preceding a sailing event. Kuteva (2001:92) states that this aspectual construction “defines a temporal phase located close before the initial boundary of the situation described by the main verb” (for a similar analysis, see König 2000:142).

These definitions show that the notion of futurity is one of the central semantic components of the prospective aspect. This notion, however, needs to be kept distinct from future tense. The difference between the two is evidenced by the fact that the prospective aspect, unlike the future tense, is compatible with past time reference (Klein 1994:116, König 2000:149). In 3, the English prospective construction to be about to V is in the simple past. The prospective event is described as being in the future relative to the subject referent’s past state.

1. (3)

   

Although one is not dealing with a (purely) temporal notion, the prospective aspect does express a temporal relation between the (current) state of the subject argument and the eventuality introduced by the main verb. I am using eventuality as a cover term for both states and events (Bach 1986). The eventuality introduced by the main verb is not realized yet, and the subject referent is in a state that is temporally anterior to this eventuality. The state preceding the prospective eventuality is sometimes referred to as a “pre-state”, as it temporally precedes another eventuality (see, among others, Ritz 2012:901, Fleischhauer & Gamerschlag 2019, Fleischhauer et al. 2019). The prospective construction says that, if nothing intervenes, the subject referent will participate in the future eventuality. The construction, however, does not state that this will definitely be the case: The example in 3 does not entail that George Canning—that is, the subject referent—actually sailed for India. The meaning of the prospective aspect is the probable inception of a future eventuality. Thus, the construction also has a modal flavor. I elaborate on this issue in section 5 (see also Bogaards & Fleischhauer, forthcoming).

German has developed at least two different periphrastic constructions for a predicative expression of the prospective aspect. The constructions in 4 and 5 represent a nominal and a verbal strategy, respectively. In both cases, the subject referent is reported to be close to exploding. The nominal strategy is to express the prospective eventuality as an NP complement of the preposition vor ‘in front of’, which itself is the complement of the (desemanticized) verb stehen ‘stand’.

1. (4)

   

The verbal strategy is to express the prospective eventuality as an infinitival complement of the pronominal adverb davor ‘in front’. Just like the vor-PP in 4, the pronominal adverb is realized as the complement of stehen. What the two strategies have in common is that the referent of the subject argument of stehen is also a participant of the prospective eventuality. In 5, it is the subject referent that is close to exploding (for a more detailed discussion of the verbal strategy, see Boogards & Fleischhauer, forthcoming).

1. (5)

   

For the current paper, I adopt the analysis proposed in Fleischhauer & Gamerschlag 2019 and Fleischhauer et al. 2019 and treat the prospective stehen vor NP construction as a light verb construction (LVC). The reason is that this construction type shares properties with other LVCs. Fleischhauer & Gamerschlag (2019:141, 146) and Fleischhauer et al. (2019:81) propose the following definition of the prospective LVC: The subject referent is close to the change of state denoted by N, which is the noun occurring within the construction (Explosion ‘explosion’ in 4). Although this is not stated explicitly, they restrict N to change-of-state (CoS) nouns, which is evident from the definition. Based on this definition, the following initial hypothesis can be formulated concerning the lexical restrictions on the type of eventuality denoted by the noun within the prospective LVC:

1. (6)

   

Change of State (CoS) hypothesis: The expression of prospective aspect is restricted to CoS expressions.

Although the CoS hypothesis is not explicitly defended within the literature, it is empirically verifiable and can be a good starting point for a corpus study of this construction. The hypothesis even gains some initial plausibility based on the existence of minimal pairs such as those in 7. Whereas the noun Verbluten lit. ‘bleeding to death’ is attested within the prospective stehen vor NP construction, I have found no single attestation of Bluten ‘bleeding’ occurring within the construction. Presented to native speakers, the sentence in 7b—without any further context—has been evaluated as unacceptable. The difference between the two nouns in 7 is that Verbluten denotes a change of state—that is, the change from being alive to being dead, but Bluten does not.

1. (7)

   

Similar restrictions are argued for with respect to prospective aspect constructions in other languages as well. According to Van Rompaey et al. (2015: 237), English prospective aspect constructions such as be on the/one’s way/road to are found in combination with transitional predicates, which express a change of state. Romaine (1999:332) reports that the prospective construction in the English-based Creole language Tok Pisin is restricted (at least with animate subject referents) to a few change-of-state verbs. König (2000:146) reports that Maa (Nilo-Saharan) shows a related albeit different restriction on one of its prospective aspect markers. All of these observations are captured by the CoS hypothesis in 6.

However, despite the fact that the CoS hypothesis gains some initial support, I argue that it is too strong. The corpus study presented in the next section does indeed show that the prospective stehen vor NP construction has a clear preference for CoS nouns; yet other semantic types are possible in the construction as well, as long as they are licensed by context. Van Rompay et al. (2015) report similar results for the English prospective be on the/one’s way/road to construction. However, the authors propose that in this case, a transitional interpretation is imposed on the nontransitional predicate. For state predicates, such as being able to in 8, for instance, they argue that the verb is converted into a dynamic predicate meaning ‘acquiring the ability to do something’.

1. (8)

   

I argue in section 5—contra Van Rompay et al.—that state nouns and process nouns are not shifted toward a transitional interpretation when used within the construction. Rather, the transitional meaning is contributed by the prospective construction itself.

3. Corpus Study: Prospective Stehen vor NP Construction.

Before turning to the details of the corpus study of the prospective stehen vor NP construction, different uses of the sequence stehen vor NP should be discussed. In particular, it is important to distinguish between the prospective interpretation (which is the focus of this paper) and the literal one. The two meanings of the construction are the topic of section 3.1. The corpus study is reported in section 3.2.

3.1. Prospective and Nonprospective Interpretations of Stehen vor NP.

A corpus study of the prospective stehen vor NP construction needs to take into account that the prospective reading is just one interpretation of the construction. Among other interpretations, stehen vor NP can have a literal interpretation as well, as shown in 9 (in section 3.2, I discuss another nonliteral interpretation of the construction). In its literal—or heavy—use, stehen is a posture verb that expresses that its subject referent is located at some spatial location in an upright posture. The verb contributes the posture information, whereas the exact spatial location of the subject referent is indicated by the verb’s PP complement, as in 9.

1. (9)

   

In contrast, in the prospective construction, stehen is semantically reduced.² The sentence in 10 does not make a statement either about the subject referent’s spatial location or about its posture. Rather, the main predicational content is contributed by the PP complement. The construction locates the subject referent of stehen in a temporal prephase of the event of bankruptcy. Dropping the PP forces a heavy interpretation of stehen, which affects the interpretation of the entire predication. Without the vor-PP, the sentence can only mean that the clinic building was still standing in 1996.

1. (10)

   

The prospective interpretation of 10 results from a metaphorical interpretation of the originally spatial preposition vor ‘in front of’. In its spatial interpretation, vor designates a spatial region in front of an entity (Wunderlich & Herweg 1991:778). The conceptual metaphor time is space (for example, Boroditsky 2000) maps spatial relations onto temporal relations. As a result, vor comes to designate a phase preceding the eventuality denoted by the PP-internal noun (Fleischhauer & Gamerschlag 2019:147). The prospective interpretation depends on the vor-PP alone. For example, in 11, vor der Explosion lit. ‘in front of the explosion’ alone has a prospective interpretation.

1. (11)

   

As mentioned above, Fleischhauer & Gamerschlag (2019) and Fleischhauer et al. (2019) analyze stehen as a light verb, and the prospective stehen vor NP construction as an LVC (German: Funktionsverbgefüge). Light verbs are semantically reduced and are frequently used for expressing aspectual notions such as stativity, inchoation or causation (for example, von Polenz 1987). As a light verb, stehen fulfills two functions in the prospective LVC. First, it licenses the vor-PP as its complement (just as the heavy verb stehen licenses the spatial PP complement). Second, it introduces a state predication (following Fleischhauer & Gamerschlag 2019:148, in examples such as 11, this state predication is expressed by the light verb stehen). As already mentioned in the introduction, prospective aspect expresses a relation between a current state and a prospective future eventuality. The prospective future eventuality is expressed by the vor-PP, while the state is contributed by the light verb (see Fleischhauer & Gamerschlag 2019 and Fleischhauer et al. 2019 for a semantically compositional analysis of prospective stehen vor LVCs).

The discussion in this section of different interpretations of the stehen vor NP sequence as well as the brief consideration of LVCs provide the relevant background for the corpus study of the prospective construction presented in the next section. In the remainder, I use the terms prospective LVC and prospective stehen vor NP construction interchangeably.

3.2. Identifying Prospective LVCs in the Corpus.

To establish whether the prospective LVCs are indeed restricted to CoS nouns, it is necessary to identify the (semantic) range of PP-internal nouns. To achieve this aim, I searched the German Reference Corpus (Deutsches Referenzkorpus, DeReKo; Leibniz-Institut 2021), which consists of more than 50.6 billion words, using the online search tool COSMAS II (Leibniz-Institut für Deutsche Sprache 2020). The search was carried out in the written language archive (Archiv W), which contains written texts of different genres, both printed and from the internet (for example, Wikipedia, chats).

LVCs cannot be easily identified within corpora since the two components, the light verb and the PP, are not necessarily adjacent but can be separated by various elements (Fleischhauer 2021:64). For example, in 12, the PP vor der Explosion is separated from the light verb stehen by the subject NP der Motor ‘the engine’ and the two adverbial modifiers allerdings ‘although’ and kurz ‘short’.

1. (12)

   

The only restriction is that the light verb and the PP occur within the same sentence. Therefore, the search was carried out with the use of the string &stehen /s0 vor, which resulted in 1,241,876 hits.³ DeReKo internally restricts the extraction of hits to 10,000 sentences. For the current study, the maximal number of sentences was extracted and manually annotated by the author and a second native speaker. Nineteen sentences were incomplete and therefore excluded from the annotation process. Once the annotation procedure was set up as described below, the annotation process was completed independently by the two annotators.⁴ After all the sentences were annotated, the results were compared. In case of disagreement, a third annotator was consulted to reach a majority consensus for each sentence.

The annotation procedure consisted of four steps. The first three steps are outlined in this section. They were used to identify prospective LVCs within the entire pool of tokens. The fourth annotation step applied to PP-internal nouns within prospective LVCs only and is discussed in section 4.1. In the first annotation step, the grammatical status of the vor-PP was annotated. Since the vor-PP is a complement of both heavy and light stehen, it was important to distinguish adjunct PPs (such as in 11 above) from complement PPs. The relevant criterion separating adjuncts from complements is whether the PP can be left out without affecting the interpretation of the verb or the acceptability of the sentence. In 13a, the vor-PP cannot be left out without resulting in an ungrammatical sentence; the PP is therefore classified as a complement of stehen. This is different for the vor-PP in 13b, which is used as a temporal adjunct and can be left out without affecting grammaticality. Examples such as 13b are not relevant for the present analysis.

1. (13)

   

The results of the first annotation step are summarized in table 1. The 3,870 sentences in which the vor-PP is not realized as the complement of stehen were excluded from further analysis.

Table 1. Results of the first annotation step: grammatical status of the vor-PP.

In the second step, it was established whether stehen was used as a heavy verb or as a light verb. The criterion applied in this case was whether stehen could be substituted by a verb from the same semantic class. When used as a light verb, stehen cannot be replaced by another posture verb, such as sitzen ‘sit’ or liegen ‘lie’, as shown in 14.

1. (14)

   

The verb stehen also has a purely locational use (Kaufmann 1995:111), as shown in 13a. In this use, there is no reference to the subject referent’s upright posture. To identify such locational constructions, the verb stehen was substituted with purely locational predicates such as positioniert sein ‘be positioned’ and lokalisiert sein ‘be localized’. For example, sentence 13a allows stehen to be replaced by one of the purely locational predicates, as shown in 15a. In contrast, sentence 14 does not allow such replacement, as shown in 15b.

1. (15)

   

The results of the second annotation step are summarized in table 2.

Table 2. Results of the second annotation step: heavy versus nonheavy use of stehen ‘stand’.

The third annotation step consisted in identifying the prospective LVCs among nonheavy uses of stehen, as not all of them have a prospective interpretation. Prospective LVCs can be paraphrased as “the subject referent is close to the eventuality denoted by N (if nothing intervenes)”. In contrast, LVCs such as vor dem Problem stehen ‘face a problem’, as in 16, cannot. This LVC is better paraphrased as “be confronted with the problem”.

1. (16)

   

Fleischhauer & Gamerschlag (2019:141) term examples such as 16 “challenge LVCs”. Prospective LVCs and challenge LVCs can be differentiated on the basis of the three criteria summarized in table 3.

Table 3. Criteria to differentiate prospective LVCs from challenge LVCs.

First, only prospective LVCs license a temporal interpretation of spatial modifiers such as kurz ‘short’. In example 12 above, kurz specifies the time interval between the current state and the onset of the prospective event. Second, only prospective LVCs allow the substitution of stehen by light sein ‘be.’ As shown in 17a, stehen and sein can both be used in prospective LVCs. However, substituting stehen by sein in a challenge LVC results in an ungrammatical sentence, as shown in 17b (repeated 16).

1. (17)

   

Third, challenge LVCs can be causativized by the light verb stellen ‘put’, as illustrated by the pair of examples in 18.

1. (18)

   

In contrast, prospective LVCs cannot be causativized at all. Thus, there is no causative version of, for instance, vor der Explosion stehen ‘be on the verge of explosion’ meaning ‘to cause something to be on the verge of explosion.’ The sentence Sie stellten den Motor vor die Explosion lit. ‘They put the motor in front of the explosion’ can only have a literal interpretation but not a prospective one.

The combination of the three criteria allows us to identify 1,606 sentences containing a prospective LVC. The results appear in table 4.

Table 4. Results of the third annotation step prospective paraphrase.

As table 4 shows, the individual occurrences can be classified into 305 types, based on the PP-internal noun: Prospective LVCs that belong to each type share the same PP-internal noun. A complete list of the nouns occurring within prospective LVCs can be accessed online.⁶ Thus, in addition to identifying the prospective LVCs among constructions with nonheavy stehen, the third annotation step also yielded a list of individual PP-internal nouns within these LVCs. The majority of the remaining nonheavy uses of stehen represent instances of challenge LVCs, and the rest resist a clear classification and seem to be idiomatic constructions. As the focus of the current paper is on prospective LVCs, I do not go into further details concerning the nonprospective stehen vor NP constructions.

4. The Semantic Type of the PP-Internal Nouns.

Having identified the individual nouns occurring within prospective LVCs in the analyzed corpus sample, I now move to the fourth step and analyze the semantic type of these nouns. Recall that according to the null hypothesis, PP-internal nouns in the construction denote a change of state. To verify this hypothesis, I apply a number of tests to determine the semantic type of these nouns.

4.1. Identifying CoS Nouns.

CoS predicates, irrespective of whether they are verbal or nominal, express a change in a property of their theme argument. As such, all CoS nouns are eventive and dynamic. However, just like CoS verbs, CoS nouns vary with respect to durativity: Some of them are durative (accomplishment predicates), while others are punctual (achievement predicates). To identify nouns that belong to each category, different tests are required, as discussed below.

For the purposes of this analysis, based on Ehrich & Rapp 2000:253, I adopt the following two criteria the authors propose for identifying durative CoS nouns. First, only durative CoS nouns license time-span adverbials such as innerhalb von drei Wochen ‘within three weeks.’ In 19, the time-span adverbial indicates that the result state—being ruined—has been achieved after a time span of three months.

1. (19)

   

Second, only durative CoS nouns can be modified by incremental adjectives such as allmählich/schrittweise ‘gradual’ (for a similar analysis of the English adverb gradually, see Piñón 2000). In 20, the noun Ruin ‘ruin’ is modified by the incremental adjective allmählich, which results in the interpretation that the ruining of the house proceeded incrementally.

1. (20)

   

Unlike durative CoS nouns, punctual CoS nouns can be identified on the basis of their compatibility with time-point adverbials such as in der 80. Minute ‘in the 80th minute’ or um Mitternacht ‘at midnight.’ Time-point adverbials go best with achievement nouns but are considerably worse in combination with durative CoS nouns or non-CoS nouns (such as process nouns or state nouns; Asher 1993:23). For example, reporting on a soccer game, it is possible to say Der Ausgleich passierte in der 80 Minute ‘The tie happened in the 80th minute’, which qualifies Ausgleich ‘tie’ as an achievement noun.

In contrast, Bluten ‘bleeding’ is a process noun, and as such is incompatible with a time-point adverbial. The sentence Das Bluten der Kuh passierte/ereignete sich um Mitternacht ‘The bleeding of the cow happened at midnight’ is odd. In combination with the durative CoS noun Verbluten ‘bleeding to death’, um Mitternacht ‘at midnight’ indicates the time at which the result state was attained and thereby focusses on the punctual final culmination.

As the fourth annotation step, the above-mentioned criteria were applied to nouns occurring in the PP-internal position of the prospective LVCs. The results of the annotation are summarized in table 5. Roughly 80% of the nouns are classified as denoting a change of state, which shows a clear preference for such nouns.⁷ These results weaken the CoS hypothesis significantly, since roughly one fifth of the nouns do not belong to this type.

Table 5. Results of the fourth annotation step CoS nouns

Among the most frequent non-CoS nouns are Spiel ‘game’ (29), Krieg ‘war’ (11), and Title ‘title’ (6); the nouns Rente ‘retirement’, Rekord ‘record’, and Unterschrift ‘signature’ appear 3 times each.⁸ In the remainder of this section, I concentrate on the non-CoS nouns, which account for almost 20% of the sample, and suggest a semantic classification for them. The two relevant semantic features underlying the classification are i) eventiveness and ii) stativity.

4.2. Eventiveness.

As argued in section 3.1, the main predicational content of prospective LVCs is contributed by the PP-internal noun. It is reasonable to suppose that the construction selects for an eventive noun in PP-internal position.

To identify eventive nouns among non-CoS nouns two criteria were used. First, only eventive nouns can be targeted by temporal and aspectual modifiers such as gestrig ‘yesterday’s’, um fünf Uhr ‘at five o’clock’, and andauernd ‘continuous’. Second, only eventive nouns can be realized as the subject of a verb indicating a temporal duration, such as dauern ‘last’ (Fábregas et al. 2012:170f.). It is possible to say that a process, such as a game, or a state, such as marriage, lasts for a certain period of time, as shown in 21a. In contrast, one cannot say that a title has a temporal duration, as shown in 21b. In fact, nouns such as Titel ‘title’ do not denote an eventuality at all; for the purposes of this study, I refer to such nouns as “noneventive nouns”.

1. (21)

   

Punctual nouns denote events without any significant duration. Thus, they do not easily combine with dauern ‘last’ but can be realized as the subject argument of geschehen ‘happen’—which is a criterion for dynamicity—and combine with time point modifiers, as already mentioned in the section 4.1. The results are presented in table 6.

Table 6. Noneventive PP-internal nouns in prospective LVCs.

As the table shows, 15 nouns within the sample do not denote an eventuality. They are listed along with their individual token frequency (N). The remaining non-CoS nouns are eventive and are discussed further in section 4.3.

4.3. Stativity.

The second relevant semantic feature for a classification of non-CoS nouns is stativity, which distinguishes event-denoting nouns from state-denoting ones. The term process noun refers to event-denoting nouns which do not express a change of state, whereas the term state noun refers to state-denoting nouns. There exist a number of criteria for (nominal) stativity (see Fábregas & Márin 2012:38f. and Fábregas et al. 2012:172ff. for a detailed discussion of the different tests), but for reasons of space, I only use one test. Process nouns can be realized as the subject of predicates such as geschehen/passieren ‘happen’, beenden ‘stop/finish’, and unterbrochen sein ‘be interrupted’, as in 22a, while state nouns cannot, as shown in 22b (Fábregas et al. 2012:170, Fleischhauer 2021:66).

1. (22)

   

On the basis of this test, 13 nouns have been classified as state nouns and 31 nouns have been identified as process nouns. The state nouns are listed in table 7, and the process nouns are listed in table 8.

Table 7. State nouns occurring PP-internally in the LVCs of the prospective family.

Table 8. Process nouns occurring PP-internally in the LVCs of the prospective family.

To summarize, the discussion in this section has revealed that four different semantic types of nouns are attested within the prospective LVCs. The relative frequency of the different types within the sample is presented in table 9.

Table 9. Relative frequency of the different semantic noun types within the sample.

These data show that the hypothesis that the prospective LVCs are restricted to CoS nouns must be rejected, if the restriction is interpreted as applying to the semantic type of PP-internal nouns. However, it may be suggested that this is a restriction on the type of eventuality that nouns denote when they appear in the construction. This would allow for mechanisms such as coercion to affect the interpretation of non-CoS nouns. Such a view has been expressed for the English prospective be on the way construction by Van Rompay et al. (2015), as I already mentioned in section 2. The authors propose that non-CoS predicates are shifted toward a transitional interpretation. I discuss this issue in the next section and propose a distinction between the transitional meaning contributed by the construction and the meaning of the PP-internal noun.

5. Interpretation of the Prospective Stehen vor LVCs.

In discussing how the interpretation of the prospective LVCs emerges, one needs to distinguish between the meaning of the light verb construction and the meaning of the embedded nominal predicate. The prospective construction has an inceptive meaning, which can be illustrated by an example such as 23. The construction contains the noun Abschluss ‘completion’, which denotes a change of state from not completed to completed. The prospective construction entails that the jobs are not completed yet, as one cannot say #Die Arbeiten stehen unmittelbar vor dem Abschluss, sie sind sogar schon abgeschlossen ‘The jobs are close to completion; in fact, they are already completed’. This indicates that the event has not culminated yet, although the process that would result in completion might have already started. The construction in 23 expresses a probable future inception of a change of state, that is, if nothing intervenes, there will be a transition from non-N (incompletion) to N (completion).

1. (23)

   

Note that the inceptive interpretation of the construction is independent of the semantic type of the noun. With state nouns, the inceptive interpretation of the construction results in a CoS reading of the noun. A representative example of a prospective LVC containing a state noun is shown in 24. The subject referent is close to the state denoted by the noun Rente ‘retirement’. As the state does not hold at the reference time, the inception of the state constitutes a change of state. This does not mean that the state noun undergoes a semantic shift and becomes a CoS noun; instead, the inceptive interpretation of the construction triggers the CoS reading of the noun: The change from being not retired to retired results from the inceptive meaning contributed by the construction.

1. (24)

   

When a prospective LVC contains a process noun, the inceptive meaning of the construction does not result in a CoS reading of the noun. For example, the construction in 25 contains the process noun Bürgerkrieg ‘civil war’, and it is presupposed that the war has not yet started at the reference time; otherwise, the use of the prospective construction would be infelicitous. Yet there is still a change of state involved: Although Bürgerkrieg does not entail the attainment of a specific result state, it entails that the eventuality denoted by its antonym Frieden ‘peace’ ends. Krieg and Frieden form a pair of antonyms (for example, Davies 2012), and the beginning of one eventuality entails the ending of the other.

1. (25)

   

It should be pointed out that this does not hold for all types of process nouns. For example, Spiel ‘game’, the most frequent process noun within the analyzed sample, does not have an antonym. I return to this issue in section 6.2, as it plays a central role in determining current relevance of such eventualities. What is important for the moment, however, is that the construction focusses on the onset of the process. What the prospective construction in 25 expresses is not the probable future attainment of a specific result state, but the probable onset—or inchoation—of a future process. This is also supported by the fact that 25 can be paraphrased as “being close to the beginning of a civil war”.

As discussed in sections 4.1 and 4.2, the construction also licenses noneventive nouns, such as Titel ‘title’. Since the prospective eventuality is contributed by the PP-internal noun, nouns such as Titel ‘title’ need to be coerced to an eventive interpretation, as shown in 26.

1. (26)

   

A reasonable interpretation of the sentence is that the subject referent is close to receiving the title. In this particular case, the title is received upon, for instance, winning either the league or a decisive game. Thus, the sentence can be paraphrased as “the subject referent is close to winning the title”. Crucially, winning a title refers to a CoS event, which gives rise to an inferred CoS interpretation.

A comparison of the example in 26 with the one in 27 reveals that the type of inferred eventuality depends on the concrete (contextual) interpretation of the noun. In 27, Titel ‘title’ refers to a doctoral degree, which is not obtained by winning a game but by successfully completing one’s doctoral studies.

1. (27)

   

In the case of noneventive nouns such as Titel ‘title’, one is dealing with context-dependent inferences about an eventuality. This is clearly an instance of coercion as it involves “a shift from a predication involving an argument that is not an eventuality to a predication involving an argument of eventuality type” (Asher 2011:214). This does not mean that Titel becomes an eventive noun; rather, the coercion process “force[s] a predication over an eventuality that is related in some way to the denotation of the syntactically given argument” (Asher 2011:214). Depending on the context, different eventualities—that is, winning a game or completing one’s doctoral studies—are inferred. To be able to make the right inference, speakers need some reliable cues, since there might be different types of eventualities in which the referent of the noun (for instance, Doktortitel ‘doctoral degree’) can participate. I assume that the meaning of the prospective construction imposes restrictions on the type of eventuality that may be inferred, but I admit that this idea needs to be worked out in more details in future studies.

At the end of section 4.3, I stated that the CoS hypothesis can be rescued if change of state is understood as referring not to the semantic type of the PP-internal noun but to its interpretation. Under this view, non-CoS nouns are acceptable in the construction as long as they can have a CoS reading—either through pragmatic inference or through coercion. As argued in this section, coercion—that is, a shift between semantic types—only happens in the case of nouns that do not refer to an eventuality; neither state nouns nor process nouns undergo coercion. The reason is that the prospective preposition vor needs an eventuality as its complement. Since state and process nouns denote eventualities, there is no type conflict. Coercion is therefore not required. In contrast, noneventive nouns such as Titel do not denote eventualities, which yields a type conflict that is resolved by coercion (as described above).

To summarize the discussion so far, the prospective LVCs have an inceptive meaning: They express the inception of a prominent part of the eventuality denoted by the nominal predicate. In the case of state and CoS nouns, as well as noneventive nouns after coercion, the prominent part is the onset of the (resulting) state; in the case of process nouns, the prominent part is the onset of the process denoted by this noun. In any case, the construction locates the subject referent in a state temporally preceding the appropriate transition.

6. Current Relevance.

Although I have shown that the prospective LVCs do not impose any lexical restrictions on the PP-internal noun, not all nouns are equally acceptable within this construction. Acceptability varies on two levels: i) semantic type and ii) meaning of individual nouns. On the level of semantic type, the construction shows a clear preference for CoS nouns. This raises the question of why such a preference exists. On the level of individual nouns, I demonstrate that some nouns require more contextual support than others when used within a prospective LVC. The preference for a specific semantic type and the need for contextual support can both be accounted for in terms of the notion current relevance.

The analysis presented below builds on Fleischman’s (1982, 1983) claim that the function of prospective aspect is to portray a future eventuality as being of current relevance (for example, for the subject referent). She argues that there is a relationship between perfect (or retrospective) aspect and prospective aspect as both establish a link between some (past or future) eventuality and a present state. More specifically, both types of aspect convey that the past/future eventuality is of current relevance to the present.

6.1. Current Relevance and Perfect Aspect.

Current relevance is a central concept in the analysis of the perfect aspect. However, it is a vague notion and, as Depraetere (1998:599) states, it is usually either defined as “continuance of the result of the past event into the present” (Dahl & Hedin 2000:391) or not defined at all. I do not aim at articulating a precise definition of current relevance; instead, I discuss this notion in the context of my analysis of three types of nouns presented in the next section.

The literature on the perfect aspect offers different definitions of the perfect, for instance, “resultative perfect”, “experiential perfect”, and “hot news perfect” (see Comrie 1976, Ritz 2012, among others for a discussion of the different takes on the perfect, and especially Thieroff 1992, chapter 6.5 for German data). The different views of the perfect aspect gave rise to different analyses, with the two most prominent being the current relevance approach and the “extended now” approach (see, for instance, McCoard 1978).⁹ The current relevance approach, on which I focus in this study, is usually formulated with reference to the so-called resultative perfect, as shown in examples such as Peter has arrived (and he is still there). The perfect in this sentence—as opposed to the simple past—gives rise to the implicature that the result state (that is, Peter being here) still holds at the reference time (see Anderson 1982:232, Thieroff 1992:175). The continuance of the result state into the present is a pragmatic inference licensed by the perfect aspect.

Dahl & Hedin (2000) propose that current relevance is both a graded notion and a pragmatic concept (see Schaden 2013 for a similar view). Although a continuant result is the most salient way in which a past event can be of current relevance, it is not the only one. Imagine that John is at a party and is supposed to be picked up by Peter. If someone says “Peter has arrived”, Peter’s arrival is relevant as it signals that John is about to leave the party. Thus, as far as the speaker is concerned, it is not the continuance of the result state that is relevant at the reference time (Peter is here), but the consequences of this result state (John is leaving the party). A different example is provided by Carey (1995). She states that the sentence John has mowed the lawn so he can come to the movies indicates “that the past event (the completion of the lawn mowing) is related to John’s present ability to go to the movies” (p. 83). In this case, the relevant inference is explicitly stated.

Yet the prototypical view of the notion current relevance is in terms of the continuance of a result, as mentioned above. This explains why there is a strong preference for telic predicates in resultative perfect constructions. However, it is obvious that current relevance may not have the same interpretation when it comes to the connection between present and future eventualities. Since a future eventuality has not been actualized yet, its prospective result cannot continue into the present moment. This means that the notion current relevance requires different approaches depending on whether it applies to a past or to a future eventuality.

6.2. Current Relevance and Prospective Aspect.

In this section, I analyze three types of nouns that occur in prospective LVCs: CoS nouns, process nouns that have antonyms (such as Krieg ‘war’), and process nouns that do not have antonyms (such as Spiel ‘game’). Based on this analysis, I propose a definition of current relevance that makes it applicable to both, perfect and prospective constructions.

The first type of noun to be discussed here denotes a change of state. This is the most frequent type of noun within the analyzed data. As discussed above, at first glance, the notion current relevance is not applicable to prospective constructions. Unlike perfect constructions, which express the continuation of the result of a past event into the present, prospective constructions describe events that have not yet happened, and so their results may not hold at the reference time. To extend the notion current relevance to prospective constructions, I propose that in this case, what is relevant for the present is not the result of a past event but the future change of state. Prospective constructions with CoS nouns describe a future event in which the subject referent undergoes a change of state. Intuitively, this change definitely has current relevance for the subject referent, as it will affect him or her directly. However, if the referent is inanimate, as in 28 (repeated 23), it seems strange to say that the prospective change has any current relevance.

1. (28)

   

The prospective completion of the remaining smaller jobs is definitely not relevant for the jobs themselves. However, the context of the example suggests that the completion is relevant for other event participants. The sentence following the one in 28 says that it would allow one to move on to the next step in the development project. Thus, the prospective change of state is of current relevance for the project managers and the investors who participate in the development project.

Under this approach, in case of prospective LVCs, what has current relevance is the prospective change of state. This change may be perceived as relevant either for the subject referent or for other event participants who benefit in some way from the change affecting the subject referent. Note, however, that in either case, current relevance is predicated upon a prospective attainment of some result state. I use the term prospective result to capture the idea that it represents the mirror image of the continuous result expressed by perfect constructions.

Based on the data that contain prospective LVCs, the most common interpretation of current relevance is in terms of prospective result. The preponderance of this interpretation explains the clear preference of prospective LVCs for CoS nouns. As these nouns lexically specify a result, they provide a natural basis for this interpretation.

The second type of noun addressed in this study denote processes that are logically incompatible with some other eventuality. The beginning of the process denoted by such a noun entails the end of the opposite eventuality. As discussed in section 5, the noun Krieg ‘war’ has an antonym, Frieden ‘peace’. In 29, Krieg appears in a prospective LVC: Two countries—Russia and Turkey—are about to go to war with each other.

1. (29)

   

The event denoted by the noun war has two or more participants. One can think of different ways in which a prospective war might be of current relevance to them. For example, a physical war—as opposed to, for instance, a trade war—is a violent event that causes damage and suffering.

However, the noun Krieg does not specify a result state, and it is possible for two enemies to declare war but not end up fighting at the end. Nonetheless, current relevance of the prospective result is applicable in such examples as well. I propose that in this case, the prospective result is achieved by ending the eventuality denoted by the noun’s antonym—in the case of 29, peace.

Finally, I have also analyzed process nouns that do not have antonyms, such as Spiel ‘game’. A result-based interpretation of current relevance seems less obvious for such nouns: They do not lexically specify a result state; neither do they denote a process whose beginning entails the ending of some other eventuality. Within the analyzed sample, Spiel is the most frequently occurring noun from this category. As a starting point, consider the sentence in 30.

1. (30)

   

If this sentence is presented without any context, native speakers are unsure as to whether or not to accept it. However, as soon as one provides an appropriate context, the acceptability of the sentence increases. Such context can be provided, for instance, by common knowledge that the Brazilian team is one of the best teams in the world, and that playing against the Brazilians is either very challenging (for the team) or very exciting (for the team and/or the fans). To be able to interpret this sentence, the listener needs a context that would make this particular event of gaming relevant for its participants.

Additional evidence for the view that examples such as 30 require contextual support to gain current relevance is provided by the fact that within the analyzed sample, Spiel is always accompanied by an attributive modifier. These modifiers frequently express an evaluation of the prospective event, as shown in 31 and 32. The sentence in 31 reports on a football game, and the modifier conveys the importance of the game for the subject referent. It is clear from the context that the subject referent is a football team, which is facing the finals in the national football league for the very first time.

1. (31)

   

In 32, the event is evaluated as challenging for the subject referent, which is conveyed by the modifier schwer ‘difficult’. From the context of this sentence, it becomes clear that the team is about to play against teams that so far have been more successful in the ongoing season.

1. (32)

   

Unlike the evaluative interpretation in 30, which is based on context alone, the evaluative interpretation in 32 arises based on the overt modifier. However, in both examples, the prospective event is being evaluated. I propose that by evaluating the prospective event as challenging or important the speaker signals current relevance for its participants.

Besides modifiers that convey the difficulty degree or the importance of an event, the noun Spiel also occurs with enumerating modifiers such as erst- ‘first’, fünfzigst- ‘fiftieth’ or Final- ‘final’. Enumerative modifiers mark the number of occurrences of an event. This could be the first or the last occurrence of an event in a series that include a specified number of such events—for instance, the first/last game of a season. Alternatively, this can be a one-off occurrence that would count as important in a given context—for instance, an anniversary. An example is shown in 33: The next game will be the 50th game in the coach’s career, which is explicitly referred to as an anniversary.

1. (33)

   

“Having completed 50 games as a coach” is a new property of the subject referent gained after the competition of the prospective gaming event. Although Spiel is not a CoS noun, the enumerating modifier gives rise to a prospective result interpretation of the prospective LVC.

Note that changing the modifier in 33 may affect the interpretation of the prospective result. For example, if fünfzigsten ‘fiftieth’ is replaced with letzten ‘last’, the sentence would mean that after the game the subject referent will no longer be the coach with the German Bundesliga. Furthermore, using a different modifier may cancel the prospective result interpretation all together. For example, using wichtigsten ‘most important’ rather than fünfzigsten ‘fiftieth’ would give rise to an evaluative interpretation of the prospective LVC. The interpretation would be that the prospective game is the most important one in the subject referent’s career as a Bundesliga coach. For example, it might decide whether this would be his first victory as the coach.

6.3. Ways to Convey Current Relevance.

I am now in a position to provide a tentative answer to the question of how a prospective eventuality gains current relevance. Based on the analysis of three types of nouns presented in section 6.2, I have identified two ways in which potential future eventualities may become relevant for an eventuality participant. First, a prospective eventuality is of current relevance if there is some prospective result associated with that eventuality. The result may be inferred based on the lexical semantics of the noun, the inceptive meaning of the construction (as with noneventive and state nouns) or derived from the context. If the result interpretation arises from lexical meaning, it can be based either on the lexically specified result (CoS nouns) or on the entailed change from the eventuality denoted by the noun’s antonym to the eventuality denoted by the noun (nouns such as Krieg ‘war’). This analysis provides a natural explanation for the high frequency of CoS nouns, since such nouns give rise to the prospective result interpretation due to their lexical meaning. Second, a prospective result of some eventuality is of current relevance if it is evaluated as important or challenging for the participants of that eventuality.

The discussion of the noun Spiel in the previous section is particularly revealing as it highlights the role of context in establishing current relevance. It has been demonstrated that there exist (lexical) cues that trigger the current relevance interpretation. In particular, adnominal modifiers allow inferring a prospective result or—more often—trigger an evaluative interpretation of the prospective eventuality. Such cues are more important in case of nouns such as Spiel than CoS nouns, since the latter license a prospective result interpretation by virtue of their lexical meaning. In case of nouns such as Spiel, a result inference does not arise on the basis of their lexical meaning alone. Although the discussion of process nouns has been restricted to Krieg and Spiel, they are representative of a broader category of nouns showing a similar behavior. For example, one of the nouns in the sample that are similar to Krieg is Unruhe ‘riot’, whereas nouns similar to Spiel include Partie ‘game’ and Einsatz ‘use’. What is especially remarkable is that all of the Spiel-type nouns in the sample occur with the same adnominal modifiers.

Finally, the discussion has shown that current relevance is not necessarily judged from the perspective of the subject referent of the prospective LVC. Rather, a prospective eventuality can be of current relevance to some other individual who might somehow be affected by it.

Before closing this section, I come back to the examples in 7 repeated below as 34.

1. (34)

   

As mentioned in section 2, Verbluten ‘bleeding to death’ is attested within prospective LVCs, but the noun Bluten ‘bleeding’ is not. Verbluten denotes a change of state, and so its occurrence in the construction is not surprising, especially as it expresses the potential death of the subject referent. In contrast, the example in 34b is infelicitous, as expected. The noun Bluten denotes a process. Without an appropriate context, the future event of the victim bleeding does not necessarily have any current relevance. The noun does not entail the attainment of a result. Therefore, a prospective result interpretation is not available for 34b. Without any supporting context, bleeding is also neither particularly challenging for the subject referent. However, this does not mean that such a context might not exist but it just shows that if it exists, it is hard to find.

7. Conclusion

In this paper, I took up the idea that prospective aspect is the mirror image of the perfect aspect. Following the proposal in Fleischman 1982, 1983, I have extended the notion current relevance, which is central to many analyses of the perfect, to prospective aspect. A current relevance approach to prospective LVCs provides an answer to a number of questions. First, prospective LVCs show a strong preference for CoS nouns: By virtue of their meaning, these nouns trigger a prospective result interpretation of the construction, which seems to be the dominant way of making the eventuality they denote currently relevant. Second, some nouns require more contextual support when used in a prospective LVC than others. In particular, the noun Spiel ‘game’ and similar process nouns require a specific context or a linguistic cue to establish the current relevance of the eventuality they denote. Adnominal modifiers have been identified as particularly useful cues for guiding the search for current relevance. The use of such modifiers is possible because the eventive predicate (or the predicate from which the eventuality is inferred) is realized as a noun.

The current paper presents the first attempt to explore the notion current relevance in relation to the prospective aspect. Pending questions for future research include i) Are there any other ways to interpret the notion current relevance (of a prospective eventuality)? Is it possible to formalize this notion, for example, along the lines of Schaden’s (2013) proposal? ii) Is it possible to identify other cues that would facilitate the interpretation of a particular eventuality as currently relevant? iii) Can the results reported in the present paper be extended to other prospective aspect constructions in German as well as in other languages?