In languages such as European Portuguese (EP), the so-called periphrastic constructions (cf. Barroso, 1994, 1997)—typically made up of a semi-auxiliary, a prepositional-like connective¹ 1

The precise category to be assigned to this type of element has been the subject of heated debate in syntactic studies, particularly whether they are true prepositions, complementizers or a hybrid category. As this paper is essentially concerned with semantic issues, I will not comment on the question here. In this respect, see, for example, Martins (2000), Oliveira et al. (2004), or Magro (2005).

When they function as aspectual operators—such as the progressive estar a (literally, ‘be at’), começar a (‘begin/start’) or para de (‘stop’), their semantic properties are relatively stable across contexts of utterance and their meanings do not change considerably.

Nevertheless, some of these constructions often show a certain variability regarding their semantic properties. For instance, the structure acabar de may convey either purely aspectual information, roughly corresponding to the English aspectualiser “finish,” as in (1), or a mainly temporal meaning, leading to an interpretation in which the situation has come about just before the speech time, as the translation of (2) makes clear (cf. Cunha, 1998; Oliveira et al., 2004).

(example taken from https://relogiodagua.pt/produto/retrato-do-artista-quando-jovem/).

This semantic diversity also seems to be present in the structure estar para (literally ‘be to’) + infinitive, which will be the central topic of analysis in our paper. In fact, although it usually conveys a pre-preparatory state leading to a given event, this construction, depending on its context of occurrence, may be associated with other interpretations, sometimes quite different from the prototypical one.

In effect, linguistic factors such as the tense form with which it is combined, the presence of some temporal adverbials or the embedding under negation play an important role in the several readings ascribed to the structure estar para, affecting the computation of its final meaning.

The main goal of this paper is, thus, to investigate the semantic properties that characterise the structure estar para + infinitive, seeking to recognise the linguistic factors that interfere with the various meanings associated with this construction and to establish the relationships between the different readings that can be observed.

The work is structured as follows: in Section 2, I will demonstrate that, in its basic reading, the structure estar para + infinitive describes a pre-preparatory state, as defined, e.g., in Moens (1987); Section 3 discusses the role played by tense in the interpretation of the construction at hand, with particular focus on the contrast between presente do indicativo (simple present), pretérito imperfeito (imperfective past), pretérito perfeito (terminative past) and future tenses. Section 4 deals with the interpretative consequences of embedding estar para under negation. In Section 5, I will look more closely at some apparent exceptions, and in 6, I will try to give a unified analysis for the semantics of the construction at issue. Finally, 7 summarises our findings.

In its most natural reading, the structure estar para + infinitive points to a period of time that immediately precedes the happening of a given eventuality, leading to its occurrence, as the sentences in (3) and (4) suggest:

In (3), the foreign investment did not yet reach Portugal at the interval of the utterance, but its arrival is taken as granted and imminent. Similarly, in (4), the release of the CD set has not yet taken place but is described as being very close in time.

Examples like these show that estar para + infinitive typically describes an interval that immediately precedes the occurrence of a given situation, serving as its onset.

One possibility to deal with this meaning would be to consider the structure estar para + infinitive as a case of “prospective aspect,” as defined in Klein (1994). According to the author, temporal and aspectual relations may be modelled by exploiting three different time intervals: the time of utterance (TU), the time of the situation (TSit), and the topic time (TT), the time about which the speaker or writer makes a claim or assertion, i.e., the interval that is focalised in discourse.

Within Klein’s framework, the prospective aspect is conceived as follows: the TT coincides with TU, and both precede TSit ((TT = TU) < TSit). Applying this definition to the structure estar para + infinitive, we obtain the desired result, according to which the construction at hand focalises an interval that is simultaneous to the speech time and immediately goes before the occurrence of the situation in its domain.² 2

The concept of prospective aspect has been widely employed in the literature to describe the linguistic behaviour displayed by different structures. One example is the use of this proposal to account for the “be going to” construction in English (cf. Schroeder, 2011) and its counterpart (ir a + infinitive) in Spanish (cf. Bravo Martín, 2008; Moreno Burgos, 2013, 2014).

Nevertheless, if we take a closer look at some examples, we observe that this merely topological definition is insufficient to account for the behaviour and restrictions exhibited by estar para + infinitive.

First, it is unable to explain some important aspectual constraints that characterise the construction at issue. Consider the following sentences, in which I indicate the basic aspectual class of the eventuality that is combined with estar para + infinitive:

Apart from a few exceptions that I will address in Section 5, it is clear that estar para + infinitive combines nicely with every sort of event but gives rise to semantic anomaly when it occurs with states. Such a disparity in terms of its selection restrictions cannot be captured through a treatment exclusively based on relationships between time intervals, as this does not satisfactorily account for the aspectual constraints mentioned above. In particular, provided that the only prerequisite associated with the prospective aspect, as formulated in Klein (1994), is the anteriority relation of TT/TU with respect to TSit, why should states, but not events, be ruled out in these contexts?

Another shortcoming of choice for merely temporal-based approaches to account for the structure estar para + infinitive has to do with the clear stative behaviour exhibited by this construction. As an illustration, consider the interpretation of temporal adverbials such as “há N tempo” (literally, “there is N time”) in the context of the presente do indicativo (simple present).

In languages like European Portuguese, the adverbial “há N tempo” receives different readings depending on whether it occurs in the simple past³ 3

With the Pretérito Perfeito (the terminative simple past), the adverbial “há N tempo” merely locates a situation within a past interval that is itself placed N time before the speech time, meaning something like “N time ago”; this reading is irrelevant for our discussion, since it does not distinguish between states and events, so I will disregard it in my analysis.

Now, as can be seen in (16)–(18), even in combination with basic event predicates, the construction estar para + infinitive cooccurs without any problems with the adverbial “há N tempo,” which is given a durative interpretation, thus suggesting that we are dealing with true stative constructions.

In addition to the behaviour brought out by the adverbial “há N tempo,” there are many other arguments that lead us to conclude that the sentences in which the structure estar para appears are stative.

First, much like lexical states (19) and differently from events (20), the structure estar para + infinitive does not typically promote quantificational or habitual readings when combined with the presente do indicativo (simple present), getting instead a purely temporal interpretation (cf. (21)):

On the other hand, much like lexical states, and in sharp contrast with eventive predications, the construction estar para + infinitive never occurs embedded under aspectualisers such as começar a (‘begin’) or parar de (‘stop’), as the following sentences illustrate:

A third argument that settles the stative nature of the construction estar para + infinitive has to do with its behaviour in the context of some triggers of agentivity (cf. the discussion in Dowty, 1979). Like most lexical states, the structure estar para + infinitive avoids agentive contexts, even when applying to basic agentive predicates: it cannot occur as an imperative (cf. (25)) and typically rejects the embedding under verbs of command, like forçar (‘force’) or persuadir (‘persuade’) (cf. (26)).

The data we’ve just discussed shows that the structure estar para + infinitive not only provides temporal information, namely by establishing a relation of anteriority of the topic time with respect to the time of the situation, but also plays an important role at the aspectual level, insofar as, on the one hand, it establishes restrictions on the nature of the predications it can be combined with, and, on the other, it shows a clearly stative behaviour.

In order to accommodate these observations, we could extend to the treatment of estar para + infinitive the analysis that Schroeder (2011) proposes to account for the construction “be going to” in English.

According to Schroeder, the prospective aspect can be defined in terms of the present relevance of a future situation, thus making explicit the temporal condition associated with estar para + infinitive. To ensure the relevance of the present domain, she assumes the existence of a state, simultaneous with the speech time, which occurs prior to the relevant eventuality, which she calls the causative state.

To that extent, the prospective aspect would be represented by a causative state, simultaneous to the utterance time, and a target state that would correspond to the situation predicted to occur in the future.

Although, in general terms, this proposal seems to be perfectly adequate for describing the behaviour of the structure estar para + infinitive, it nevertheless raises some questions of undeniable importance that need to be solved.

On the one hand, the state that precedes the occurrence of the relevant situation need not be a causative one, as the following example suggests:

In a sentence like (27), it seems hard to find an explicit causal chain that would originate the future occurrence of the situation described by “the book coming out.” On the contrary, we are dealing with a state that merely precedes the occurrence of the event at stake.

On the other hand, as we have seen above, the eventuality that follows the state that overlaps the speech time is never a stative predicate but almost always an event, so the use of the term “target state” to describe it seems quite inappropriate.

In order to avoid these shortcomings, I will choose here to adopt a treatment based on the phase structure proposed, e.g., by Moens (1987) or Moens and Steedman (1988), that seems to be more suitable to correspond to the behaviour of the construction at hand.

Moens (1987) and Moens and Steedman (1988) conceive the distinct types of events as the result of the combination of different phases that constitute the so-called Aspectual Nucleus. Except for statives, which do not exhibit a complex internal constitution, each aspectual class is, thus, characterised by its phrasal structure, obtained from different patterns of interaction between a preparatory process, a culmination point, and a consequent state.

Besides this nuclear makeup, it is also possible to refer to a phase that occurs immediately before the beginning of a given eventuality. Though intimately related to it, this phase cannot be considered a portion of the nuclear structure of the situation at hand since it does not take part, by itself, in the internal structure characterising the event (cf. Moens, 1987; Moens & Steedman, 1988; Binnick, 1991; Kamp & Reyle, 1993). Since this phase is clearly stative in nature, it is called a pre-preparatory state.

Serving to refer to the coming about of a given eventuality, the pre-preparatory state does not seem to significantly interfere in the internal constitution of the situations with which it is combined, although it has a relevant impact at the aspectual level.

Treating estar para + infinitive as a pre-preparatory state has the advantage of accounting for the various restrictions manifested by its behaviour, contrary to the purely temporal proposal by Klein (1994), which is unable to deal with the aspectual constraints that we have observed. On the other hand, being more neutral regarding the nature of the phases that are involved, this treatment also avoids the shortcomings that were recognised in the analysis developed by Schroeder (2011).

In sum, following Moens (1987) and Moens and Steedman (1988), I will argue for an analysis of the construction estar para + infinitive, at least in its standard use, as the linguistic exponent, in EP, of a pre-preparatory phase, conceived as an optional state that precedes the nuclear phase structure of a given event.

Given that basic states do not have their own phase structure, i.e., they do not take part in the aspectual network. This analysis correctly predicts that examples like (8) and (9), repeated here for convenience as (28) and (29), are excluded since phases cannot be added to eventualities that do not present a complex internal constitution.

On the other hand, assuming that pre-preparatory phases are states of their own right explains the stative behaviour associated with the structure estar para + infinitive.

As we have already mentioned, the construction estar para + infinitive, when combined with the presente do indicativo (simple present), introduces a pre-preparatory state that is simultaneous with the speech time, and that leads to the future realisation of a given event, as illustrated in the following examples:

Thus, in a sentence like (30), the structure estar para + infinitive refers to a period of time – overlapping the utterance time - that precedes the occurrence of the event corresponding to the “approving a law on the control of arms sales”, a situation that is presented as certain in the future.

Similarly, in (32), the birth of the goat’s kids is not only assumed to occur in a future time but, due to the expression of a pre-preparatory phase carried out by estar para + infinitive, it is also taken as relevant at the present interval, thus explaining the “soften of the loss of the four ewes.”

However, when we combine estar para + infinitive with past tense forms, namely the pretérito imperfeito do indicativo (the imperfective EP past tense), and the pretérito perfeito do indicativo (the terminative EP past tense), the results are quite different.

Beginning with the pretérito imperfeito (imperfective past), we observe that the structure estar para + infinitive combined with this tense form describes an ongoing pre-preparatory state located in the past but very often carrying an inference of non-realisation of the accompanying event, as shown in (33)–(34):

We observe that in a sentence like (33), the event that follows the pre-preparatory state, viz. the return to Portugal, is not meant to come about; the same happens with the situation corresponding to the plan going to the Council of Ministers in (34).

Note, nevertheless, that this interpretation is just an inference (most likely an implicature) associated with the use of the imperfeito (imperfect) since, in many other sentences, the actual occurrence of the eventuality at issue seems to be confirmed. (35) and (36) illustrate this point:

The events that follow the (past) pre-preparatory state in these examples, namely writing in (35) and being born in (36), really took place at a past interval, which proves that the non-realisation of the situation associated with estar para + infinitive in the imperfect is undoubtedly a cancellable inference.

In any case, the behaviour of the structure estar para + infinitive with the pretérito imperfeito (past imperfect) contrasts sharply with what happens when the configuration at hand appears in the present tense, insofar as, in the former case, and given the appropriate conditions, the occurrence of the situation that follows the pre-preparatory state may not be guaranteed, which brings this construction close to some modal structures that convey uncertainty.

The move from a purely aspectual meaning into a modal one is even more evident when we consider the combination of the structure estar para + infinitive with the pretérito perfeito (the EP terminative past tense).

Consider the following examples:

It is clear that both the discussion of the subject at the beginning of the meeting in (37) and the marriage in (38) never took place; moreover, unlike the pretérito imperfeito (imperfect), the pretérito perfeito (terminative past) never admits the realisation of the event that follows the pre-preparatory state, thus functioning, in a certain extent, as a structure that refers to counterfactual propositions.

Thus, the inference of non-realisation of the event following the pre-preparatory state seems to be part of the basic meaning of the construction estar para + infinitive in the pretérito perfeito, as the clear-cut contradiction shown in the following example demonstrates:

The discussion developed so far leads us to the following conclusions: (i) the choice of different tense forms has a decisive influence on the final interpretation of the construction estar para + infinitive; the use of past tenses such as the pretérito imperfeito (imperfective past) and the pretérito perfeito (terminative past) favours a move from a purely aspectual interpretation, i.e., prospective aspect, into clearly modal meanings – uncertainty readings in the case of the imperfect, counterfactual ones in the case of the terminative past.⁶ 6

This move from prospective aspect into irrealis modality brought about by the use of past tenses in contrast with the simple present was already noticed and discussed for languages like Turkish, Basque, and Papuan by Jendraschek (2014), who claims that, diachronically, while a present prospective can develop into a future tense, past prospectives tend to take over the function of counterfactual conditionals in the past.

This move towards a modal meaning has an interesting consequence: unlike its counterpart with the present tense, the structure estar para + infinitive in the pretérito perfeito (terminative past tense) may be easily combined with basic statives. Consider the following pair of sentences:

Ultimately, the contrast in acceptability that emerges between (40) and (41) reflects the fact that the aspectual restrictions we observed for the presente do indicativo (simple present) do not extend to the pretérito perfeito (terminative past), which would confirm our hypothesis that, actually, in the latter case we are dealing with a modal operator and not with an aspectual one.

Not surprisingly, we did not find in our corpus the combination of the structure estar para + infinitive with future tenses. This may be explained by the well-known fact that, in the majority of the languages of the world, the expression of posteriority in the future is highly restricted, if not forbidden (cf. Comrie, 1985; Binnick, 1991; Declerck, 1991, 2006).

The only possibility available is the combination of estar para + infinitive with the futuro simples do indicativo (simple future) in its epistemic modal meaning, as illustrated in (42)–(43):

In these examples, we are not dealing with a posteriority-in-the-future reading, but rather with an epistemic interpretation, in which the futuro simples (simple future) is used to emphasize the uncertainty of the speaker regarding the realisation of the event following the preliminary state.

Although, in a few cases, the embedding of the structure estar para + infinitive under sentential negation does not substantially affect its prime semantic properties, simply bringing about a reading in which the occurrence of a given pre-preparatory state leading to a future event is denied (cf. (42)–(43)), there are nevertheless examples in which we find clear changes to the basic meaning that was ascribed to the construction at hand.

As shown in the corresponding English translations, in (44) estar para + infinitive describes a pre-preparatory state leading to the future breaking up of the music group UHF, but due to the semantics of negation, this state is presented as not going on at the speech time. Similarly, in (45), the combination of estar para + infinitive with sentential negation points to an interpretation in which the preliminary phase conducting to the readjustment of the market is not occurring yet at the utterance time.

However, in the vast majority of constructions of this kind found in our corpus, the obtained results go far beyond the interpretation that reflects the mere compositional combination between the basic semantics of the construction estar para + infinitive and negation. Take the following illustrative sentences:

In examples like these, we are no longer talking about the expression of a pre-preparatory state that precedes the occurrence of a future event and which is denied, but rather about an attitude of rejection towards the realisation of the situation on the part of the subject of the sentence. Thus, in (47), for instance, the most natural interpretation is that the community does not want to support the scientists; a parallel reading can be found in (48), which points to the people’s rejection of watching live versions with a reduced cast, and in (46) regarding the event of buying shoes.

So, once again, we are dealing with a move from a pure aspectual sense into a modal meaning. In this case, negation seems to favour a reinterpretation of the construction estar para + infinitive, imposing a reading that conveys a volitional or desiderative modality.⁷ 7

For an in-depth characterisation and discussion on the concept of volitional or desiderative modality, see, among others, Palmer (2001), Portner (2009), Oliveira and Mendes (2013).

That negation plays a central role in this interpretative change can be confirmed by the following contrasts:

While the affirmative sentences (49) and (51) describe, as expected, a pre-preparatory state leading to the realisation of the events of Maria moving house and the guests arriving at the party, respectively, in their most natural readings (49) and (51) behave quite differently. In (49), we get a volitional or desiderative interpretation which expresses Maria’s desire not to move house; in (51), however, since the eventuality of arriving at the party is not agent-oriented, and, as a consequence, not subject to volitions, the sentence sounds somewhat odd.

We are therefore observing an interesting behavioural pattern concerning the structure estar para + infinitive: when combined with certain linguistic elements, this construction moves from conveying aspect to expressing modality. The modal values it acquires, however, differ greatly depending on the linguistic forms it interacts with: with the pretérito imperfeito (imperfect), it receives an epistemic interpretation of uncertainty; with the pretérito perfeito (terminative past), it discloses a reading of counterfactuality and, finally, with negation, it refers to the expression of volitional or desiderative modality.

In addition to the cases we have just discussed, the structure estar para + infinitive appears in other contexts in which its canonical interpretation seems to undergo quite visible changes, even if they are not as significant as the previous ones. In the present section, I will address some of these problematic instances.

A first kind of example, which would seemingly contradict the proposal we have been endorsing, is shown in (53)–(54):

As we mentioned above, the structure estar para + infinitive in the presente do indicativo (simple present tense) does not typically combine with stative predications. So, how can we explain these examples?

One hypothesis to consider is that, in these cases, we are not dealing with true states but rather with culminations. To that extent, “to be a mother” would not refer to the state that describes a family relationship but rather to a momentary situation equivalent to becoming a mother or giving birth. This reading is widely attested in our corpus, even in examples where the construction estar para + infinitive does not appear, as the following sentences prove:

The temporal adverbial “pouco depois das 13h00 do dia seguinte” (“just after 1pm the following day”) that locates the eventuality “ser mãe” (“become a mother”) in (55) and the quantificational adverbials “nove vezes” (“nine times”) and “onze vezes” (“eleven times”) counting the situations of being a mother in (56) clearly show that, in these sentences, we are dealing with events, not with statives.

This alternation between a stative and an eventive (more precisely, a culmination) reading is not surprising: several authors, such as Dowty (1979), have pointed out that some verbs appear in both stative and culminative predications, depending on their context of occurrence (consider, for instance, the contrast between “Mary knows French”, a state, and “Mary knew the truth yesterday,” a culmination).⁸ 8

On the similarities between states and culminations, see, among others, Verkuyl (1989) or Rothstein (2008).

More complex are the cases in which the structure estar para + infinitive denotes a durative situation overlapping the utterance time and describing the progress of a given predication. This interpretation is extremely restricted and only occurs with certain fixed or idiomatic constructions, such as “está para durar” (literally, “be to last”). Consider the following illustrative examples:

In examples like these, the structure estar para + infinitive merely describes a durative state conceived as the continuation or extension (in the present and in the future) of a given situation associated with the noun phrase in the sentential subject position. In our corpus, only “estar para durar” (literally, “be to last”) and “estar para ficar” (“literally, be to stay”; cf. (59)–(60)) can express this kind of interpretation.⁹ 9

Confirming the rather crystallised nature of this kind of interpretation associated with estar para + infinitive, we can invoke the use of a very curious expression, “lavar e durar” (“wash and last”), which obviously cannot be interpreted in its literal sense:

(i) A polémica em torno da indústria da pedra em Sesimbra, na zona do Parque Natural da Arrábida, está para lavar e durar. (CetemPúblico, par=ext509741-soc-93b-1)

The controversy in surrounding ofthe industry ofthe stone in Sesimbra, inthe area ofthe Park Natural ofthe Arrábida be.pres3sg to wash.inf and last.inf

‘The controversy surrounding the stone industry in Sesimbra, in the area of the Arrábida Natural Park, is set to last.’

Taking into account the divergent interpretations displayed by the structure estar para + infinitive in the various contexts of occurrence considered in the previous discussion, the question arises as to which semantic properties actually qualify to describe its behaviour. Can we ultimately find a unified characterisation for this construction? I will devote to this issue in the next section of this paper.

As we have seen, in addition to its canonical interpretation, in which it describes a pre-preparatory state leading to the realisation of a given event (cf. Section 2), the structure estar para + infinitive can, under appropriate conditions, convey modal information of uncertainty, counterfactuality (cf. Section 3), or desire/intention (cf. Section 4), as well as expressing a durative state that extends from the present into the future (cf. Section 5).

However, it is essential to ask if there are, in fact, semantic properties common to these divergent readings that would allow us to provide a unified characterisation for the construction at hand. We believe that the answer to this question is positive. Let us see to what extent.

Firstly, in all these different readings, the structure estar para + infinitive behaves consistently as a state, irrespective of the basic aspectual class of the eventuality with which it combines.

Thus, even when it is associated with a culmination, the construction under analysis is always durative, regardless of the specific reading it receives, as the following examples show:

In all these sentences, the durative expressions – “há meia hora” (“for half an hour”) in (61)–(62), “durante meia hora” (“for half an hour”) in (63), “durante a tarde” (“in the afternoon”) in (64) and “durante os próximos anos” (“for the next few years”) in (65)—measure out the situation associated with estar para + infinitive, proving that, irrespective of the aspectual class of the eventuality in its scope, this construction gives rise to predications that persist in time.

Moreover, estar para + infinitive, in its different readings, much like lexical states,¹¹ 11

For a more detailed discussion on the interaction of stative predications with different types of aspectualisers in European Portuguese, see, e.g., Cunha (2004) or Cunha (2011).

This kind of incompatibility seems to be primarily due to the stative nature of the structure estar para + infinitive since if we reverse the order of the operators—and taking into account that the construction under analysis typically applies to events –, the results are perfectly acceptable, as (69)–(72) reveal. In other words, we cannot justify the ungrammaticality of examples like (66)–(68) simply through the idea that there is a complete mismatch between the elements involved; the explanation will, therefore, lie in the stative nature of the structure estar para + infinitive that, as all other statives, cannot be embedded under aspectualisers such as começar a (‘begin’) or parar de (‘stop’).

There is a second property that seems relevant when it comes to proposing a unified semantic characterisation for the different meanings associated with the structure estar para + infinitive. In fact, in all the interpretations considered, there seems to be a relationship of prospective projection between the state that overlaps a given interval of time that constitutes its temporal perspective point (cf. Kamp & Reyle, 1993) and the eventuality in its scope.

In the case of the canonical interpretation of estar para + infinitive, this semantic schema is straightforwardly applicable. Consider the following sentence:

In a sentence like this, the pre-preparatory state triggered by the form estar para overlaps the speech time, which serves as its temporal perspective point, projecting the associated event of “iniciar ali uma obra de trânsito” (“starting works of traffic there”) into a future interval.

A similar treatment, provided the consideration of some adaptations, could be extended to the cases we labelled “durative” in section 5. Consider the following illustrative example:

In this sentence, too, the state associated with estar para takes place in a period of time whose temporal perspective point is the moment of the utterance, but it is also projected into a future time interval. The difference regarding examples like (73) is that the situation described remains the same, i.e., it is the state associated with the noun phrase that occupies the subject position - in this case, the controversy surrounding corruption in national arbitration - that takes place in the present and extends into a later time.

The notion of a prospective projection also seems to be involved in cases where the structure estar para + infinitive receives modal interpretations. However, in these contexts, the situation to which the operator applies is no longer located in a future time interval but rather in a possible world different from w0 (but also subsequent to the state described).

In this example, the structure estar para + infinitive describes a pre-preparatory state located in a past interval that precedes a situation referring to the stopping of the attacks and counter-attacks, which, regardless of whether or not it comes about in the real world, is typically conceived of as being posterior to that state.

Similar remarks can be extended to the cases in which a counterfactual reading is involved:

Although, in sentences like these, the situations accompanying the form estar para never happen in the real world, the idea of prospective projection remains valid. In either case, the pre-preparatory state associated with estar para necessarily precedes the virtual realisation of the eventualities described by “the students fail” or “the ship evacuates all the foreigners.” The fundamental difference is that, in examples like these, the prospective projection is no longer made regarding a time interval but a possible world that does not coincide with the world of reference.

Finally, and regardless of the semantic consequences carried over by sentential negation -a problem we will not deal with here, as it goes far beyond the scope of this article- it should be noted that, typically, volitions and intentions, presented in examples like (78), express, by their nature, situations that are assumed to occur in the future, so the idea of prospective projection also seems perfectly appropriate to describe this kind of configurations.

In short, despite the diversity of meanings discussed so far, it was also possible to find semantic invariants that characterise the structure estar para + infinitive. Thus, in all the observed readings, we can say that the construction at issue (i) always describes a stative predication that establishes an overlapping relation with a given temporal perspective point and (ii) induces a prospective projection of the situations with which it is typically combined.

Unlike most aspectual operators in European Portuguese, which maintain their meanings unchanged regardless of the conditions in which they occur, the structure estar para + infinitive is sensitive to linguistic and contextual factors that constrain, to a certain extent, the readings it triggers.

Thus, while, prototypically, this construction conveys information about the prospective aspect, displaying a state that, in the present, precedes the realisation of a future event, factors such as the combination with certain tenses or the presence of sentential negation substantially alter the obtained interpretations.

In the case of co-occurrence with the pretérito imperfeito (past imperfect), for example, the structure estar para + infinitive favours an epistemic modal interpretation, highlighting the uncertainty regarding the actual realisation of the eventuality described.

On the other hand, the pretérito perfeito (past terminative) typically induces a counterfactual reading of the eventualities involved, inscribing them in a possible world which is definitely different from the world of reference.

Finally, sentential negation gives preference to the arising of readings concerned with volition, desire or intention.

In all these cases, we observe a clear shift from a plainly aspectual interpretation to other readings involving essentially modal meanings.

However, there seem to be some semantic invariants that run through all these divergent readings, allowing us to establish a unified characterisation for the construction estar para + infinitive. In particular, we observed that, in all the cases under analysis, estar para describes a preliminary stative situation that consistently induces what we have termed a prospective projection. Such projection is meant to place the relevant eventualities either in a future interval or in a possible world.