In this feature we ask about the order of tense, aspect, and mood markers (TAM markers) with respect to each other. To qualify for this feature, it is important that the three markers be adjacent to each other. A context which favours the combination of the TAM markers is counterfactual clauses, where the tense marker usually corresponds to a past marker, the aspect marker to an imperfective marker, and the mood marker to a future marker (used modally). An example is Creolese, where the order of the three markers is tense-aspect-mood (TAM):
Note that we consider markers to be adjacent even if lexical items such as adverbs may intervene between them (see Chapter 45 and Chapter 46 on the tightness of the link between the past and the progressive marker with the verb); what matters is that the markers are located on the same side of the verbal complex.
There are many APiCS languages to which this feature does not apply, either because they do not possess three markers, or, if they have three markers, they do not allow them to combine or to be adjacent to each other.
As in other chapters, we use “aspect” in a restricted sense, applying it only to the opposition between perfective and imperfective aspect.
This feature has played an important role in creole studies since it was one of the features of Bickerton’s language bioprogram hypothesis. According to Bickerton, creole languages display the order TMA when the three markers combine (see below our value 1), where the category ‘mood’ is defined as ‘irrealis’, referring to markers which express future, counterfactual, conditional, and similar functions (Bickerton 1980: 6). Note, however, that Bickerton was not the first linguist to make this claim. To our knowledge the first linguist was Voorhoeve (1957: 384).
We distinguish the following four values:
excl | shrd | all | ||
---|---|---|---|---|
Tense-Mood-Aspect | 11 | 1 | 12 | |
Tense-Aspect-Mood | 2 | 0 | 2 | |
Mood-Tense-Aspect | 3 | 1 | 4 | |
The feature does not apply | 56 | 0 | 56 | |
Representation: | 73 |
This box shows that the possibility of using tense, aspect, and mood markers adjacent to each other is restricted to a relatively small subset of APiCS languages (about 25%); furthermore, the box also shows that the logically possible combinations MAT, ATM, and AMT are not attested in our languages.
Value 1 (Tense-Mood-Aspect) occurs in six English-based and in six French-based languages.
Value 2 (Tense-Aspect-Mood) is found in Creolese (see example 1) and in Kinubi:
Value 3 (Mood-Tense-Aspect) is found in Santome, in Principense, and in Reunion Creole; note, however, that in Reunion Creole the confidence value is given as ‘uncertain’.
In some languages, for example Guyanais, Kinubi, or Principense, the mood marker occurring in combination with the tense and the aspect marker also functions as a future marker, but in other languages like Nengee or Krio this is not the case. In Nengee, for instance, the future marker is o and the counterfactual marker is sa, and in Krio the markers are go and fɔ. However, it looks as if there has been some diachronic change from future and counterfactual marker to only counterfactual marker, at least in Sranan (which is closely related to Nengee). In Early Sranan, sa is the only marker attested as a future marker, and it also occurs in counterfactual and similar contexts. The following example illustrates value 1 (TMA):
In modern Sranan, sa only occurs in counterfactual and similar contexts, and the future marker is go. This suggests that the grammaticalization of the verb go as a future marker and the specialization of sa as a (non-future) mood marker is a later development.
Modern Sranan is the only APiCS language that allows different orders. Example (8) illustrates value 1 (TMA) and example (9) value 3 (MTA):
Note that according to Winford & Plag (2013), example (9) was only accepted by one older informant.
As noted above, value 4 (the feature does not apply) concerns (i) languages which do not have all three markers, (ii) languages which do not allow them to combine, or (iii) languages which do not allow them to be adjacent to each other. Among this last group we find for example Papiamentu, where the future marker lo has partly retained the pre-subject position of its Portuguese adverbial etymon (logo ‘right away’) and where the past marker taba has fused with the originally progressive marker ta (note that taba is not used without ta except for the verbs ta ‘be’ and tin ‘have’, as in the following example):
In several languages, some markers are preverbal and others postverbal (see also the examples in Chapter 43 on the position of TAM markers in relation to the verb):
Among the APiCS languages, the possibility of combining tense, aspect, and mood markers adjacent to each other occurs only in creole languages. This possibility is also restricted areally: it occurs in the Atlantic, the Indian Ocean, and in East Africa. In the Atlantic, it occurs in Portuguese-based, in English-based, and in French-based creoles; in the Indian Ocean, it occurs in the three French-based creoles. Note that the Portuguese-based creoles exhibiting this feature are restricted to the Gulf of Guinea; in East Africa, this feature occurs in Kinubi, which is Arabic-based.
The results of this chapter cast some doubts on Bickerton’s claim that the TMA pattern is universal. First, the possibility of combining the tense, the aspect, and the mood marker only occurs in seventeen languages; second, these seventeen languages are areally restricted, and third, six of the seventeen languages which do allow the combination of the three markers do not have TMA, but TAM or MTA.