Journal of Language and Linguistics Volume 1 Number 3 2002 ISSN 1475 - 8989

1. Introduction

The study of conditionality has been approached from different theoretical perspectives, both diachronically and synchronically (Austin 1961, Haiman 1978, 1986, Comrie 1986, Sweetser 1990, Cutrer 1994, Dancygier and Sweetser 1996, Dancygier 1998). The work of Sweetser (1990) is specially relevant from a cognitive point of view, since it demonstrates that the occurence of conditional constructions in three different cognitive domains (content, epistemic and speech act ) reflects metaphorical mappings from physical to more abstract worlds. Furthermore, Sweetser provides a generalization, showing that these same metaphorical cognitive processes operate on other syntactic phenomena such as modal verbs in English.
This paper is a contribution to the understanding of the relations between conditionality and modality, by analysing conversational data in Brazilian Portuguese. I will claim that not only conditional constructions can undergo the same cognitive processes as modal verbs do, but that they constitute by themselves one of the expressions of the more general concept of modality, which will be treated here as related to the notion of force dynamics (Talmy 1981, 1988 and Sweetser 1990 ). Furthermore, following recent work on the cultural basis of human cognition (Tomasello 1999), the specific claim is that forces and barriers relate to the uniquely human cognitive ability to understand the intentional relations that animate beings have to the external world and the causal relations that inanimate objects and events have with one another. According to this view, language embodies particular perspectives: it is possible to place the same experiential situation into different conceptual categories for different communicative purposes.

The paper is organized as follows. In section 2, I will present a brief review of the criteria that have been proposed for the definition of modality in the literature, arguing in favour of a sociocognitivelly rooted definition. In section 3, I will present the notion of grammatical construction, discussing the integration of the meaning of the skeletal construction If p,(then) q with compositional aspects of verb form in the conditional clause. I will also argue that conditional constructions are space builders, as defined and exemplified in works such as Fauconnier (1985, 1994), Cutrer (1994), Sweetser (1996, 1997), Dancygier and Sweetser (1996) and Dancygier (1998). The main idea to be argued for is that conditional constructions function interactionally in terms of intentional forces and barriers, either taking away a potentially present social barrier or presenting a compelling force directing the subject towards an act (or some kind of reasoning).

The choice of verb form in conditional protases - specifically present (indicative) or future (subjunctive) - plays an important role in this interactional process. The kind of modality being profiled will be indicated by the verb form: the choice of the more neutral subjuntive form simultaneously takes away potentially present social barriers (dealing with face protection for topic change or for the refusal of an invitation, for example) and the choice of the present form indicates that the force being imposed is taken to be a factual one, which leaves the hearer with little option for refusing the act (or reasoning) being suggested.

2. A Sociocognitive Approach to Modality

2.1. Traditional definitions of modality

Most languages have some kind of grammatical system of modality, although they appear to be able to divide up the field in various ways. An obvious example is the existence of a system of mood in Latin (indicative, subjunctive and imperative) and the occurence of a system of modal verbs in English (will, can, must, etc), which exhibit some translational correspondence.

A number of criteria have been proposed for the definition of modality. Lyons (1977) and Palmer (1986) discuss modality in terms of notions such as non-propositionality, subjectivity and non-factuality. The problem is that such concepts are either too wide or too narrow to account for the category being defined.

The notion of non-propositionality has been traditionally linked to the speaker's attitude or opinion, as opposed to the "contents of the sentence" (Jespersen 1924, Lewis 1946, Austin 1962, Fillmore 1968). Therefore, if a proposition (which may be true or false) is subject to further qualification, this qualification represents modality. The problem with this dichotomy is that it allows the inclusion of too many cases. The sentences below are given by Palmer (1986:15) to illustrate this point:

Although the examples above are related in some way to the speaker's opinion, it is not hard to see that to include them we would have to widen the notion of modality to a point where it would become useless (since language is perspectival in nature).
If we turn to the notion of subjectivity, we will be dealing again with the grammaticalization of speaker's (subjective) opinions or attitudes. However, there are modal verbs, like CAN, which express ability, and not subjective opinion. On the other hand, it is not always easy to distinguish between a subjective opinion and an objective necessity. The sentence "You must go", for example, may indicate either a general objective necessity for going or the speaker's judgement of the situation.

Similarly, the notion of non-factuality has been discussed in terms of modality. The concept has been contrasted to categorical assertions, defined as straightforward statements of facts. Contrast the sentences below:

Analogously to other criteria discussed so far, there are problems in linking non-factuality to modality, since speech act theory makes it clear that assertions not only have propositional content, but also illocutionary force (Austin 1962). Therefore, it can be argued that assertions too are subjective, representing the speaker's point of view. Moreover, as Palmer says (1986:27), there are languages in which declarative sentences belong formally to a modal system.

The weaknesses of the criteria presented above seem to be related to the view of meaning as the relationship between words and world. From a cognitive perspective, language reflects our cognitive structuring of the world and modality (as indeed many other linguistic phenomena) can only be explained by positing a unified cognitive basis for it. In the next section, I will review some aspects of the work of Talmy and Sweetser, which provide a cognitivelly based definition for modality.

2.2. A cognitive approach to modality

The work of Talmy (1981, 1988) takes an initial step in this direction, since it suggests that the semantics of root modality is best understood in terms of force dynamics. This notion points basically to the linguistic expression of forces and barriers in general: permission represents the taking away of a potentially present barrier (e.g. may, let, allow), while obligation relates to a compelling force directing the subject to the choice of some specific act (e.g. must, ought to, have to, need to).

Taking a force dynamics account of modality, Sweetser (1990) extends Talmy's proposal by arguing that modal verbs do not have two separate unrelated senses, but rather show an extension of the basic root-sense to the epistemic domain. Therefore, the following correspondence is defined:

Sweetser's basic idea is that our reasoning processes are subject to obligations, permissions and abilities, just as our real-world actions are subject to modalities of the same sort.

This idea can be reframed by recent work on the cultural origins of human cognition, which argues in favour of a socially rooted understanding of causality (and hence to the linguistic expression of this concept). In the next section, I will present some interesting findings on evolutionary anthropology by Michael Tomasello and co-workers.

2.3. The cultural origins of human cognition

Comparing nonhuman primates and human beings cognitive skills, Tomasello (1999) presented overwhelming empirical evidence suggesting that only human beings are capable of understanding conspecifics as intentional agents like the self. Therefore, although nonhuman primates are themselves intentional and causal beings, they do not understand the world in intentional and causal terms. Primates are able to understand the antecedent-consequent relations among external events, but the understanding of intentionality and causality requires the individual to understand the mediating forces - which are not readily observable - in these external events that explain "why" a particular antecedent-consequent sequence occurs as it does. As Tomasello puts it:

Moreover, just as primate understanding of relational categories evolved in social domains to comprehend third-party social relationships, human causal understanding also evolved first in the social domain to comprehend others as intentional agents. So, the hypothesis is that the uniquely human ability to understand external events in terms of mediating intentional/causal forces emerged first in human evolution to allow individuals to predict and explain the behavior of conspecifics and has since been transported to deal with the behaviour of inert objects.

For example, Nagell, Olguin and Tomasello (1993) presented chimpanzees and two-year-old human children with a rake-like tool and an object out of reach. The subjects observed the tool being used by two demonstrators: one employed a more efficient method and the other employed a less efficient one. The result was that human children learned by imitation. They copied the method of the demonstrator in each of the two observation conditions, whereas chimpanzees just imitated the act of obtaining the object (emulation learning), but not the methods they observed. While children reproduced adult behaviour even in the case of the less efficient method, chimpanzees employed lots of different methods to reach their goals. The conclusion is that imitative learning is not a "more intelligent" learning strategy than emulation learning; its simply a more social strategy, by which individuals understand others as intentional agents with whom they can align themselves.

But how does language fit into this picture? Tomasello (1999:163) answers this question by hypothesizing that the perspectival nature of linguistic symbols, and the use of linguistic symbols in discourse interactions in which different perspectives are explicitly contrasted and shared, provide the raw material out of which the children of all cultures construct the flexible and multi-perspectival -even dialogical- cognitive representations that give human cognition its unique power. Therefore, it is not the case that each event or entity has its own true label. The use of a particular linguistic symbol implies the choice of a particular perspective: people need to communicate about many different things from many different points of view.

2.4. A sociocognitive approach to modality

Taking into account the findings on the cultural basis of human cognition presented so far, this paper proposes a sociocognitive approach to modality. The notion of force dynamics is still adopted, but forces and barriers are viewed in terms of a socially rooted understanding of causation (and not purely as a result of physical manipulation of the environment). Modality is then a cover term for all those linguistic expressions which impose an interactional force by taking some explicit perspective towards some specific event, action or reasoning process.

Therefore, modality can be coded by different linguistic expressions in a language (or cross-linguistically), such as:

In this paper, I will investigate a specific kind of adverbial clause, namely conditional constructions. I will argue that these constructions function as modality operators, which either take away a possible interactional barrier or impose an interactional force on hearers.

3. Conditionals and Cognition

3.1. Conditionals as Grammatical Constructions

Cognitive Linguistics assumes that the grammar of a language is a structured inventory of conventional linguistic units (Langacker 1987:73). These units are morphemes, words, phrases and generic schemas that describe the conventional grammatical patterns which are also used to create new sentences and phrases. Schemas are acquired through exposition to already existing expressions that instantiate them.

As conceived in this framework, grammatical constructions (schematically represented) are full linguistic entities, that integrate two ou more component structures to form a composite expression (Fillmore & Kay 1993, Goldberg 1995). Langacker (1987) puts it in the following way:

Grammatical constructions are, therefore, complex categories, which are represented in the form of schematic networks. In this view, the prepositional-phrase construction includes not only a high-level schema, but also subschemas, specific expressions, and categorizing relationships that associate these various structures.

In this paper, I will focus on conditional constructions that instantiate the general schema [IF P, (THEN) Q] in Brazilian Portuguese. This construction is instantiated by a number of subschemas, related to mood and tense selection in the protasis, as represented below:

Each of the subschemas above represents a singular pairing of form and meaning, leading to a specific kind of cognitive instruction.

The present paper analyses "a" and "b" constructions above, focusing on the sociocognitive implications of the choice between the present of the indicative and future of the subjunctive in the conditional protases.

3.2. Conditionals and Cognitive Domains

Many studies have shown that conditionals in natural languages cannot be reduced to the logical notion of material implication (Austin 1961, Haiman 1978, Comrie 1986). From a logical if-then perspective, a conditional is false, only if the antecedent is true and the consequent is false. Nevertheless, speakers of natural languages rely on considerations that go beyond truth values in order to consider conditionals as well-formed. That is why logically well-formed conditionals may sound strange in language use, as in the example below:

The point is that, in natural language conditionals, events presented in the protasis and in the apodosis are expected to be related. The unnaturalness of the example above, therefore, is due to the fact that it is hard to imagine a relation between the events described.

The kind of relation involved may be one of causality. In fact, many conditionals can be described as virtual projections of direct causal manifestations. These projections take place in three domains: content, epistemic and speech act (Sweetser 1990). Table 1 below illustrates these possibilities:

As shown above, the kind of cognitive domain in which virtual projecton of direct causal manifestation takes place is what allows the distinction among the three types of conditionals proposed by Sweetser (1990).

3.3. Conditionals and Mental Spaces

In mental space theory, language comprehension and production involve the setup of structured and interconected cognitive domains, which are independent of language. Linguistic expressions are conceived as surface manifestations of these subjacent and highly abstract constructions; sentences give partial and highly underspecificated instructions for: domain constructions; partitioning and subdivision of information in different domains; structuring of elements and relations inside domains; and the construction of connections between domains and between elements in different domains.

Discourse interpretation results from the construction of a configuration of spaces hierarchycally related and interconnected. As each sentence in discourse is processed, the configuration of spaces is dynamically activated, based on lexical and grammatical clues.

Spaces are structured by the activation of frames and schematic conceptualizations, such as Idealized Cognitive Models (Lakoff 1987). Besides that, they are structured by inference and reasoning processes.

As discourse develops, new spaces are set up as a result of clues given by space-builders, grammatical markers, or pragmatic information. Space-builders can take various forms: prepositional phrases, conectives, clauses that require complements, etc (for example, in the photo, in 1995, at school, if_____, He said that______, John believes that_______.).

As for conditionals, there is the setup of a configuration of spaces, that functions as an informational frame with deductive potential. In the present paper, I will focus on two kinds of conditional constructions that contrast in terms of tense/mood selection in the protasis, namely: present/indicative or future/subjunctive.

Let's observe the following examples:

In the example (5) above, the speaker creates an hypothetical space, assuming the factuality of the event described. Therefore, the conditional protasis is the viewpoint, from where a prediction is made (apodosis). In example (6), although the same prediction is made, the speaker adopts a neutral epistemic stance towards the event described in the protasis.

4. Modality and Conditionality

Given that, in Brazilian Portuguese, the speaker may take a factual (positive) or a hypothetical (neutral) perspective on events described in conditional protases, it seems worth investigating cognitive and interactional aspects which motivate one option or another.

The analysis of conversational data allowed the following generalizations:

I. The choice of present/indicative in the protasis sinalizes factuality in the communicative domain, as a way to impose interactional force on the following discourse situations:

II. The choice of future/subjunctive sinalizes possibility in the interactional domain, as a way to remove interactional barriers for the following discourse situations:

In the next section, I will provide evidence for the claimings above, based on examples extracted from the following two kinds of conversational data:

(i) Transcription of a 20h conversational interaction among high school teachers and a coordinator. The aim of the interaction was to have participants talking informally about their teaching practices in order to improve them.
(ii) Transcription of a 2h TV program, which has a conversational frame, involving a journalist (mediator), a government technician, a theater director and a writer.

4.1. Topic Introduction and Topic Change

According to Chafe (1994), one avaiable pattern of topic development is elicitation. In this case, there is a sequence of brief turns, involving two or more interlocutors. Topic development is achieved interactionally: the elicitator introduces the topic, which is developped by other participants.

In what follows, the analysis will focus on conversational data obtained by elicitation. It is a pedagogical meeting for the discussion of topics related to language teaching, and there is a coordinator who introduces some topics which will be discussed by a group of teachers.

In the following example, the conversation is organized around the supertopic TEXT GENRES; the topic ESSAY is elicitated by the coordinator through a conditional protasis. In this case, the verb is in the present tense of indicative mood, as shown below:

In the interaction above, the topic ("essay") is introduced by the coordinator through a conditional protasis in the present tense. However, the contrastive mention of another argumentative text ("editorial") is done through a conditional protasis in the future. It is possible to conclude, therefore, that this is a discoursive-interactional choice , and not a choice based on pre-estabilished "reality", since we have no reason to think that "essays" should be more factual than "editorials". Since, in this case, the mention of the "editorial" constitutes a contrastive temporary topic change, the secondary role it takes in the interaction is indicated by the choice of an epistemic more neutral form - the subjunctive future.

A clear example of the selection of the subjunctive future for topic change occurred on a TV debate, involving a government technician who works on social statistics (EN), a theater director (GV) and a journalist (LN):

In the conversational interaction above, one of the participants displays a conflicting view with respect to the scopus of the research on brazilian people alimentary habits being presented by a government technician. The strenghtening of the conflict is avoided by a topic change, initiated by the journalist`s use of a subjunctive future conditional.

4.2. Directives: Persuading or Requesting?

It has been shown in the literature that conditionals are particularly productive as a way of checking the felicity conditions of a speech act. Let us take some of Sweetser`s examples of speech-act conditionals to illustrate this point (1990:118):

If we now turn to Brazilian Portuguese, we might wonder which verbal form would be appropriate for the protases above: present or future? The answer seems to be related to the kind of speech act performed. Speech act conditionals are mainly directives; in the sense that their apodoses represent attempts by the speaker to get the addressee to do something. As Searle (1976) pointed out, these directives can be requests, questions, orders, invitations, suggestions or advises.

Interestingly enough, Brazilian Portuguese makes a distinction among these directives with respect to tense selection in the protasis. The future of the subjunctive occurs when the directive points to a future act of the hearer that will occur in the speaker's interest. Requests, invitations, questions and orders can be included in this case. On the other hand, the present of the indicative will be selected when the directive sinalizes a future act of the hearer that is taken to occur in the hearer's own interest, as in suggestions and advices.

The following present tense conditional was found in an advertisement run by JB FM radio station:

In the advertisment above, and in many others, the main goal is to persuade the audience to do something presented as an action that will take place in the hearer's interest. The suggestion is based on a cultural fact (it evokes the ICM for "rush hours in big cities"), which allows the choice of the present tense in the protasis.

On the other hand, future tense conditionals can be found in requests and invitations such as the one presented below:

In the examples above, the actions requested are conditionally presented, so that the speaker checks the felicity conditions for their fullfilment.

5. Conclusions

The sociocognitive approach to conditional constructions in Brazilian Portuguese brought evidence to the relation between conditionality and modality. The main conclusions that could be arrived at were the following:

1. Conditional constructions are modality operators, in the sense that they impose force or remove barriers in the interactional domain;

2. Tense-mood selection indicates the kind of force dynamics involved:

3. The removal of sociocognitive barriers through future tense allows the negotiation for topic change or the accomplishment of face threatening directives (invitations, requests and so on);

4. The imposition of sociocognitive force through present tense sinalizes topic
introduction or the accomplishment of directives intended to persuade;

6. Notes

(I) Tomasello, M. 1999. The Cultural Origins of Human Cognition. Cambridge, Massachusetts: Harvard University Press.p. 23

(II) Langacker, R. 1987. Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequisites. Stanford, Calif.: Stanford University Press.pp.409-10

7. References

Austin, J.L. 1961. Ifs and Cans. In Philosophical Papers. Oxford:Clarendon Press: 153- 80.

Austin, J.L. 1962. How To Do Things With Words. London: Oxford University Press.

Chafe, W. 1982. Integration and Involvement in Speaking, Writing and Oral Literature. In Deborah Tannen, ed., Spoken and Written Language: Exploring Orality and Literacy. Noorwood, N.J.: Ablex.

Chafe, W. 1994. Discourse, Consciousness and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago & London: The University of Chicago Press.

Comrie, B. 1986. Conditionals: a typology. In E.C. Traugott, A. ter Meulen, J. Snitzer Reilly, and C.A.Ferguson (eds.), Cambridge: Cambridge University Press

Cutrer, M. 1994.Time and Tense In Narrative And Everyday Language. Ph. D. diss., University of California, San Diego.

Dancygier, B. 1992. Two Metatextual Operators: Negation and Conditionality in Polish. Proceedings of the Eighteenth Annual Meeting of the Berkeley Linguistics Society, 61-75.

Dancygier, B. 1993. Interpreting Conditionals: Time, Knowledge and Causation. Journal of Pragmatics 19: 403-434.

Dancygier, B. 1998. Conditionals and Prediction. Cambridge: Cambridge University Press.

Dancygier, B & Sweetser, E. 1996.Conditionals, Distancing, and Alternative Spaces. In A . Goldberg, ed. Conceptual Structure, Discourse and Language, 83-98. Stanford, California: CSLI Publications.

Dancygier, B. & Sweetser, E.1997. Then in Conditional Constructions. In Cognitive Linguistics 8-2, 109-136.

Dancygier, B. & Sweetser, E. 2000. Constructions with if, since and because: Causality, epistemic stance, and clause order. In Couper-Kuhlen, E. & Kortmann, B. (eds), Cause, Condition, Concession, Contrast; Cognitive and Discourse Perspectives. Berlin/New York: Mouton de Gruyter.

Fauconnier, G. 1985. Mental Spaces: Roles and Strategies. Cambridge, MA: MIT Press.

Fauconnier, G. 1994. Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge: Cambridge University Press.

Fauconnier, G. 1996. Analogical Counterfactual. In Spaces, Worlds and Grammar. Chicago & London: The University of Chicago Press. 57- 90.

Fauconnier. G. 1997. Mappings in Thought and Language. Cambridge: Cambridge University Press.

Fauconnier, G. & Turner, M. 1994. Conceptual Projection and Middle Spaces. Technical Report no. 9401, Department of Cognitive Science, University of California, San Diego.

Fillmore, C. 1968. The Case for Case. In Bach, E. and Harms, R.T. (eds.), Universals in Linguistic Theory. New York: Holt, Rinehart, 1-88

Fillmore, C. 1990 a . Epistemic Stance and Grammatical Form in English Conditional Sentences. Papers from the Twenty-sixth Regional Meeting of the Chicago Linguistic Society, 137-162.

Fillmore, C. 1990 b . The Contribution of Linguistics to Language Understanding. In Bocaz, Aura (ed.), Proceedings of the First Symposium on Cognition, Language and Culture. Universidad de Chile, 109-128.

Fillmore, C. & Kay, P. 1993. Construction Grammar. Ms, University of California, Berkeley.

Fleischman, S. 1990. Tense and Narrativity: From Medieval Performance to Modern Fiction. Austin, Tex: University of Texas Press.

Goffman, E. 1974. Frame Analysis. New York: Harper and Row.

Goldberg, A . 1995. Constructions. Chicago: University of Chicago Press.

Haiman, J. 1978. Conditionals are Topics. Language 543, 564-589.

Haiman, J. 1986. Constraints on the form and meaning of the protasis. In E.C. Traugott, A.ter Meulen, J. Snitzer Reilly, and C. A. Ferguson (eds) On Conditionals.

Jespersen, O. 1924. The philosophy of grammar. London: Allen and Unwin.

Lakoff, G. 1987. Women, Fire and Dangerous Things: What Categories Reveal about Mind. Chicago: University of Chicago Press.

Lakoff, R. 1973. The Logic of Politeness: or minding your p's and q's. Proceedings of The Ninth Regional Meeting of the Chicago Linguistic Society, 292-305.

Labov, W. & Waletzky, J. 1967. Narrative Analysis: Oral Versions of Personal Experience. In June Helm, ed. Essays in Verbal and Visual Arts: Proceedings of the Annual Spring Meeting of the American Ethnological Society. Seattle: University of Washington Press.

Lerner, G. 1996.On the Semi-Permeable Character of Grammatical Units in Conversation: Conditional Entry into the Turn Space of Another Speaker. In E. Ochs, E. Schegloff & Thompson, S, eds. Interaction and Grammar, 238- 276. Cambridge: Cambridge University Press.

Lewis, C.I. 1946. An analysis of knowledge and evaluation. La Salle, Ill.: The Open Court Publishing Co.

Lyons, J. 1977. Semantics (2 volumes). Cambridge: Cambridge University Press.

Nagell, K., Olguin, K. & Tomasello, M. 1993. Processes of social learning in the tool use of chimpanzees (Pan troglodytes) and human children (Homo sapiens). Journal of Comparative Psychology 107, 174-186.

Palmer, F.R. 1986. Mood and Modality. Cambridge: Cambridge University Press.

Searle, J.R. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.

Searle, J.R 1976. The Classification of Illocutionary Acts. Language in Society 5, 1-24.

Silva-Corvalan, C. 1983. Tense and Aspect in Oral Spanish Narrative: Context and Meaning. Language 59:760-780.

Sweetser, E. 1990. From Etymology to Pragmatics. Cambridge: Cambridge University Press.

Sweetser, E. 1996. Mental Spaces and The Grammar of Conditional Constructions. In Sweetser & G. Fauconnier, eds, 318-333. Spaces, Worlds and Grammar. Chicago: The University of Chicago Press.

Talmy, L. 1981. Force dynamics. Paper presented at the conference on Language and Mental Imagery, May 1981, University of California at Berkeley.

Talmy, L. 1988. Force dynamics in language and cognition. Cognitive Science 2.49-100.

Traugott, E. 1985. Conditional Markers. In J. Haiman, ed., Iconicity in Syntax, 289-310. Amsterdam/Philadelphia: John Benjamins.

About the Author

Dr Ferrari lectures at Universidade Federal do Rio de Janeiro, Brazil.

Email: ferrari@west.com.br