Systems of TMA and Related Categories in Functional Discourse Grammar: A Brief Overview

Abstract In this introductory paper, we will briefly introduce FDG and its treatment of systems of tense, aspect, mood, evidentiality, polarity, mirativity and localization, as well as their lexical counterparts.


Introduction
This special issue of Open Linguistics presents a number of studies in the field of tense, mood, aspect and related categories carried out within the theoretical framework of Functional Discourse Grammar (FDG, Hengeveld & Mackenzie 2008, Keizer 2015).The related categories include evidentiality, polarity, mirativity, and localization (see Table 1 below), but also the lexical counterparts of these grammatical categories.
It is characteristic of the treatment of these categories in FDG that they are analysed as belonging to two different modules of linguistic organization, the first module concerning linguistic action and the second linguistic designation.In each module, the notions of hierarchy and scope are used to explain a number of properties of TMA systems, such as the ordering and co-occurrence or non-occurrence of the relevant morphemes, as well as the degrees of their grammaticalization.
In this introductory paper, we will briefly introduce FDG and its treatment of systems of tense, aspect, mood, evidentiality, polarity, mirativity and localization, as well as their lexical counterparts, in Section 2. In Section 3 we then show how the individual papers in this volume fit in the FDG approach to these categories.
govern morphosyntactic and phonological encoding.The scope relations are twofold: (i) the Interpersonal Level occupies a hierarchically higher position than the Representational Level, and (ii) within each of these levels there is a hierarchical organization of layers.This twofold hierarchy is represented in Figure 1.In this figure the symbols '>' when read from left to right within the two levels and '↓' when read from top to bottom mean 'has scope over'.
At the Interpersonal Level, i.e. the level of linguistic action, different layers are recognized, the scope relations between which are as follows: the lowest layer relevant here is the Ascriptive Subact (T), which represents an act of predication; the Communicated Content (C), which is the message transmitted in an utterance, scopes over the Ascriptive Subact; the Illocution (F), which captures the communicative intention of the speaker has the Communicative Content in its scope.The highest layer relevant here is the Discourse Act (A), which represents the basic unit of communicative behaviour.At the Representational Level, i.e. the level at which the relation between language and the external world is established, the lowest layer is the Property (f), which is expressed by a lexical element; next, the Configurational Property (f c ) consists of the lexical element and its argument(s) and as such provides the basic characterization of a State-of-Affairs; the Configurational Property is in the scope of the Stateof-Affairs itself, i.e. the real or hypothesized situation the speaker has in mind, which can be situated in relative time; then follows the Episode (ep), which is a thematically coherent set of States-of-Affairs that are characterized by unity or continuity of time, location, and participants and can be located independently in absolute time; the highest layer at the Representational Level is the Propositional Content (p), which is a mental construct entertained about an Episode and cannot be located in either space or time.
FDG distinguishes two types of grammatical categories: (i) relational ones and (ii) non-relational ones.The former establish a relation between non-hierarchically related entities, such as those between a verbal lexical property and its arguments.These are accounted for as functions, such as Actor and Undergoer at the layer of the Configurational Property.Non-relational grammatical categories, such as tense, mood and aspect, are treated as operators in FDG.Among these we can, again, distinguish two sets: (i) those that operate on the noun phrase, and (ii) those that operate on other entities.The former group of noun phrase operators are not dealt with in this volume.The latter group may, in other approaches, be treated as clausal operators, but this would be incompatible with the FDG approach: not all Discourse Acts are realized as clauses.For instance, the expression Why? may occur independently, and despite not being a clause it unarguably expresses the grammatical category of interrogativity, hence it is a Discourse Act.This is why the Discourse Act rather than the clause is taken to be the relevant domain of investigation.
This wide definition of operators outside the domain of the noun phrase opens up an enormous range of grammatical categories.Table 1 gives an overview of all the relevant categories as presented in Hengeveld & Mackenzie (2008), but it includes the modifications of the theory proposed in later publications (see below).
Although it is outside the scope of this introduction to give a full motivation of all the categories presented in Table 1, some comments are in order.We will start by commenting on the categories that appear at both the Interpersonal and Representational Levels, then comment on some of the level-specific categories.Polarity is generally assumed to be a category that operates on multiple levels (see e.g.Horn 1989, Dik 1997: 169-187, and Mackenzie 2009), but Hengeveld & Mackenzie (2018) show that from a typological perspective there is even reason to distinguish grammatical expressions of negation at each and every layer of the Interpersonal and Representational Levels.
With regard to the Interpersonal Level, within FDG different contradictory claims have been made as regards the position of mirativity, described in some detail by Hengeveld & Olbertz (2012) (on the typology of mirativity) and Olbertz (2012) (on the relation between mirativity and exclamativity), but given the strategic function of mirativity it can adequately be accounted for as an operator of a Communicated Content (Hengeveld 2017).
Turning to the Representational Level, it should be noted that FDG now considers mood and evidentiality to be different categories, the latter category being described in detail in Hengeveld & Hattnher (2015).With regard to mood, we need to explain the different kinds of modality distinguished in Table 1, which are based on Hattnher & Hengeveld (2016).
Proposition-oriented modality (p-modality) equals subjective epistemic modality, which, in turn, should be distinguished from objective epistemic modality, operating at the Episode-layer (ep-modality); the linguistic relevance of this distinction is discussed in detail by Olbertz & Hattnher (2018).Furthermore, episode-oriented modality contains evaluative deontic modality, which scopes over prescriptive deontic modality, one of the event-oriented modal distinctions (cf.Olbertz &Gasparini-Bastos 2013 andOlbertz &Honselaar 2017).All in all, event-oriented modal distinctions (e-modality) concern the permissibility, desirability or the circumstantial possibility of the occurrence of a State-of-Affairs.Participant-oriented modality (f c -modality), finally, concerns the permission, capability and desire of some animate individual to participate in some State-of-Affairs.
As regards aspect, the label for the event-oriented category "quantificational" is self-evident, but it should be noted that qualificational aspect, which operates at the Configurational Property, contains phasal aspect, including prospective aspect and resultative perfect, as well as the perfective-imperfective dichotomy.
In fact, there are probably few languages or even none at all that possess grammatical expressions for all the distinctions given in Table 1.However, this hierarchical structure is intended to make sense for lexical expressions of these distinctions as well.The basic structure of FDG at each layer of representation is such that the semantic distinctions given in The variable of a given layer, represented by 'v' in (1), can be specified by an operator (π).The colons in this representation should be read as "such that".The variable (v 1 ) is restricted by a (lexical) head, and possibly further restricted by a modifier (σ), which is a lexical expression, typically an adverbial one.Therefore, whatever distinction a language may be unable to express grammatically can probably be expressed by means of modifiers.Although lexical expression is of course much more variable than grammatical expression, FDG takes the stance that lexical expressions of the distinctions presented in Figure 1 function at the same layers within the same hierarchical structure as the grammatical ones.

This volume
The papers in this volume focus on different portions of the set of operators listed in Table 1.The order of presentation of the papers is consistent with the overall top-down organization of FDG, and starts with papers covering the highest level and layers of organization, working down to those dealing with lower layers of organization.There is one paper that deals with both Interpersonal and Representational distinctions, one on mirativity, one on evidentiality, two on modality, one on negation at the Representational Level and three on the perfect.
In 'A'ingae (Cofán/Kofán) operators' Kees Hengeveld and Rafael Fischer consider the overall system of TMA in this Andean language and check its compatibility with FDG predictions.The conclusion of the paper is that most A'ingae data confirm these, but that in the areas of quantificational aspect and reportativity the theory needs to be adapted.
In her paper 'Mirativity in Mandarin' Fang Hongmei argues that the sentence-final particle le in Mandarin Chinese expresses mirativity, something which had remained unnoticed in the literature thus far.She furthermore provides evidence that mirativity in Mandarin should be located at the layer of the Communicated Content at the Interpersonal Level in FDG.
Lois Kemp's paper 'English evidential -ly adverbs in main clauses: a functional approach' deals with the lexical expression of evidentiality in English.English evidential adverbs are argued to be located at various layers of both the Interpersonal and Representational Levels.The author shows that where an evidential adverb has multiple uses in English, these uses are located at contiguous layers in FDG, and never skip a layer.Evelien Keizer's paper, 'Modal adverbs in FDG: putting the theory to the test' looks into the lexical expression of modality by means of modal adverbs in English.These operate at various layers of the Representational Level.Studying data on the coordination, embedding and positioning of these adverbs, the author concludes that the classification of these adverbs in FDG is largely correct.However, she also finds that certain adverbs should be reclassified and offers suggestions to do so.
'Modality in Spanish Sign Language (LSE) revisited: a functional account' by Ventura Salazar-García refines the account of this problem presented within the framework of Functional Grammar by Herrero Blanco & Salazar García (2009).His findings provide evidence of the fact that the modal expressions in this language not only behave as predicted by FDG, but also in the same was as in languages of the spoken modality.His arguments concern the positioning of modal expression in the clause and their combinability in single clauses.
In 'Mitigating commitment through negation' María Jesús Pérez Quintero offers an account of 'negative raising' in FDG.She shows that the type of complement clause (in terms of layering at the Representational level) as well as the factual and/or presupposed nature of a complement clause determine the possibilities for having the negation in the main clause rather than the subordinate clause.She also shows how 'negative raising' can be accounted for in FDG without resorting to movement operations.
In 'Tense switching in English narratives: an FDG perspective' Carmen Portero Muñoz documents various cases in which the present perfect is used in a non-default way in various varieties of English.She concludes that the present perfect can fulfil various non-aspectual functions both at the Interpersonal and Julia Skala's 'The Present Perfect puzzle in US-American English: an FDG analysis', too, is dedicated to the present perfect in English, but focuses on the rare phenomenon of its use with tense-specifying adverbs.She finds that these uses have the common property that, in some way or another, they serve as a justification of an adjacent Discourse Act.
The paper 'The perfect in (Brazilian) Portuguese: a Functional Discourse Grammar view' by Hella Olbertz looks into the perfect in Brazilian Portuguese, which behaves markedly different from the cognates in other Romance languages in being unable to express resultativity.Based on a diachronic and synchronic description of the past and present perfect, it turns out that within the FDG approach to the grammaticalization of aspect and tense the curious case of the (Brazilian) Portuguese perfect can be adequately accounted for.