What makes you say so ? On the types of motivation in the domain of expressive competence

The reflection topic proposed by the coordinating editor of Energeia V, «Causal and final explanations in linguistics», is a challenge best defined as “philosophical” in nature. I choose to mark this qualification with quotes because I think it is necessary to suggest, beside the objective interpretation of the term, which is in keeping with the intentions of both the volume editor and the participants in the discussion forum, the existence of a non-technical evaluative interpretation of which we are all equally (or should I say: painfully?) aware – one in which the label “philosophical” is used in order to dismiss from the start such areas of inquiry, by suggesting that they lie outside the scope of linguistic disciplines proper. In the domain of the linguistics of languages, more precisely in historical linguistics – in investigations which deal in a direct way with issues such as linguistic change and language varieties, it is inevitable to at least touch upon the issue of causality vs. finality (purposiveness), even when not fully delving into it from a theoretical standpoint. Thus, while a concern with this issue tends to be more readily accepted in the context of studies on the diachronic description of particular languages, by contrast we rarely – if ever – see it tackled directly and explicitly in the domain of text linguistics. Nevertheless, as I will try to illustrate below (section 1.2.), the choice between causally-oriented views and finally-oriented views, most commonly instantiated in the way researchers conceive the speaking subject’s ‘freedom’ and concrete role in the process of discourse construction and interpretation, surreptitiously infiltrates studies in this domain, like a disembodied spectre unpleasantly ‘there’, but only

surfacing through 'signs' such as the underlying metaphors of the terminology employed, or seemingly innocent formulations (purportedly 'unbiased' by any "philosophical" view) which aim to interpret 'the facts'.
All this, indeed, with good reason.This problematics does have something of a ghost-like quality, in that it is always 'there', as the autochthonous inhabitant and rightful owner of a territory that we -linguists in the narrow sense so favoured today by our institutions -have only recently claimed and taken possession of.Beyond -or perhaps before -down-to-earth choices as to conceptual framework and research methodology, (friendly) controversies between linguists of different persuasions always come down to the fundamental views on language and man -or, more aptly, as Coseriu put it, "on man and his language" -that each of us entertains.The stakes are always high, because the issue of causality and (/vs.)finality is inextricable from that of our ground epistemological stance. 1  I cannot claim any qualification to pursue the debate in the realm of philosophy proper (the philosophy of language included).Instead, the goal of the present contribution to this debate is modest.I will restrict myself to the topic of the speaker's motivation(s) as can be found at the level of his linguistic competence as an intuitive or "technical" knowledge, possibly with its zone of intersection with (or, better: its first step in a transition towards) the level of reflexive knowledge.Throughout the paper, the perspective will be that of text linguistics -more precisely, of the discipline defined by Eugenio Coseriu as "text linguistics as a linguistics of sense".In other words, the only issue I intend to address here is: What kind of explanations, causal or finalistic (purpose-oriented), can a properly grounded text-linguistic approach offer as concerns the speakers' motivation(s) in constructing their discourse?Can text linguistics offer any valuable explanations at all in this respect, different from those given by adjacent disciplines that focus on individual speech from other angles (psychological, cognitive, pragmatic, sociological etc.)?
To clarify even further, I have to state that, from the whole area of a text linguistics conceived on the basis of Coseriu's conceptual framework, I intend to focus here mainly on the individual dimension of sense-construction/interpretation, without dealing specifically with the universal and the historical dimensions of texts, 2 although reference to these two latter dimesions will have to be made in the course of the discussion.Consequently, the considerations that follow are to be situated in that sub-discipline of a Coserian text linguistics which concerns itself with the explicitation and justification of sense construction processes in particular texts, i.e. to "text hermeneutics" in a restricted acceptation, as clarified by Loureda (2006: 149), with reference to the sub-field originally indicated by Coseriu through the term "stylistics of texts".

Starting point:
The nature of expressive competence 1.1.Text/discourse as a functionally autonomous level of speech -the genuine object of text linguistics as a linguistics of sense -is defined, in Coseriu's triadic model of the universal, idiomatic and particular levels of speech, through a specific type of content ("sense"), a 1 A peremptory demonstration is offered by Johannes Kabatek in his study "Intuición y empirismo" (2012).
2 For the investigation of the universal and the historical dimensions of texts, see the framework presented and developed in Kabatek 2005aand 2005b, Loureda Lamas 2006: 127-151, López Serena 2012.specific competence ("expressive competence") and specific evaluations through the value of "adequacy". 3These distinctions are summarized in Table 1.
Table 1: Coseriu's triadic model: Levels and forms of language, with their associated evaluations In this theoretical perspective, expressive competence (as in fact linguistic competence on all levels and with all its components) is understood as a technical knowledge (cognitio clara confusa and cognitio clara distincta inadaequata): a type of "intuitive" knowledge that is clear and certain, but is not justifiable or is only partially justifiable by the speakers themselves as speakers. 4Cornel Vîlcu (2002Vîlcu ( -2003) ) poignantly sums up Coseriu's view on the speakers' knowledge of language as a technical knowledge in the following way: "the speaker knows how to 'make' language, without knowing precisely why, in this situation or that, he does as he does". 5 More specifically, were the speaker asked why he chose to speak in a certain fashion under certain circumstances, as part of his own linguistic knowledge, i.e. at the level of his own competence as a speaker, he has only two possible types of "immediate motivation" 3 See Coseriu 1955-56: 285-287 and1988: 59, 70-75.For a subcategorization of adequacy (Angemessenheit) judgements, see 1981: 41-43 and 1988: 179-181.The main distinctions that inform this general outlook are already announced in Coseriu 1948, andfully developed in Coseriu 1958.For the different types of linguistic content, see also Coseriu 1969 and1970. 4 See esp.Coseriu 1968, Coseriu 1988, sections 3.2.(Die Sprachkompetenz als intuitives Wissen) and 3.3.(Das sprachliche Wissen als technisches Wissen), Coseriu 1992: 11.For a clarifying discussion, see also Kabatek 2003: 254-256. 5 "vorbitorul ştie cum să 'facă' limbaj, fără a şti cu precizie de ce, în cutare sau cutare situaţie, face astfel."(Vîlcu 2002(Vîlcu -2003: 153): 153).All English translations of quoted sequences are mine (E.T.-M.), unless otherwise specified.("unmittelbare Begründung"): 6 (a) a historical motivation, understood as recourse to a tradition of speech ('I say so because this is how we say this in English / in this community / in this situation etc.'), and (b) a functional motivation ('I say so because saying otherwise would mean something else').Any attempt at further motivation ("die Begründung der Begründung") would actually take us into the realm of reflexive knowledge, where the opinions of naïve speakers turned 'amateur linguists' no longer represent a reliable starting-point for scientific explanation.It is then the task of the linguist to provide a grounded explanation, or a justification at the level of epistemic knowledge, of the speaker's technical knowledge manifested in the activity of speaking as such. 7This is the fundamental stance that defines any linguistic (sub-)discipline as "realistic": 8 […] une discipline rationaliste et réaliste (ce qui, au fond, est la même chose) […] correspond -ou aspire à correspondre -à la réalité des faits linguistiques eux-mêmes et à l'intuition des sujets parlants (non pas à l'intuition qu'ils peuvent manifester de façon explicite, mais à l'intuition qu'ils manifestent dans leur activité même de parler et de comprendre).(Coseriu 1983a: 148) 1.2.If accepted and assumed as a basis and general chart for linguistic investigation, Coseriu's table of fundamental distinctions, simple and straightforward as it may appear, has truly momentous implications.One of the most crucial is, if I am allowed to use a spatial metaphor, the point of departure and the expected point of arrival of the researcher's sweeping regard upon his object of investigation.
Let me take a roundabout way concerning this matter, and begin by discussing an example taken from the sub-field of text typology, more precisely from the linguistic approach to text genres.
In an interesting study whose main focus is to describe and analyze the phenomenon of the "fictionalization of genres" (Coutinho / Miranda 2009), 9 the authors, who adhere to the definition of genres as abstract patterns of generic features, propose to address up-front the obvious difficulties that encumber this type of view: The relation between genre as an abstract category and text as an empirical object that is always an example of a certain genre raises epistemological and methodological questions.(Coutinho / Miranda 2009: 35) What remains undisputed throughout their otherwise commendable enterprise is the very premise of this definition, stated repeatedly in no ambiguous terms: "genres […] are un-ontological entities" (p.35) "the genre does not have ontological reality" (p.39) "the genres do not have their own ontological reality " (p. 41) This, of course, results in a genuine dilemma: 10 6 Coseriu 1988: 221-222.For a more detailed discussion, see infra, 2. 7 See Coseriu 1973Coseriu /1981Coseriu : 40-43, 1978Coseriu : 11, 1992: 11: 11.For a succinct discussion on this issue, see Kabatek 2012, esp. 102-103.[…] admitting, as we have been doing, that the genre does not have ontological reality, the components of genre would also be, necessarily, textual components.(Coutinho / Miranda 2009: 39) In other words, the genre does not have "ontological reality", but the features that define it only exist in "empirical texts", and thus, by way of consequence, do have "ontological reality".Or, in reverse, the "components of genre" are empirically "real", whereas the combinations thereof (genres) are not.Thus, although (abstract) genre and (empirical) text are reciprocally autonomous entities situated on totally different planes of 'existence', the components of the former can only be conceived, in fact, as components of the latter.
Attempting to break the apparent vicious circle, the authors propose, starting from the framework of Jean-Paul Bronckart's socio-discursive interactionism, a model of analysis that can be applied, they claim, "to both individual empirical texts and to abstract genres", by first identifying "the foreseeable characteristics that constitute the identity of the genre", called "parameters of genre" (p.40).These abstract parameters are presumed to be subsequently "assumed" by each text and "fixed as empirical texts" through "mechanisms of textual realization" (p.41).
On the path from abstract genres to empirical texts, these "mechanisms" induce a twofold degree of mutability or flexibility: (a) on the one hand, parameters of genre are taken out of their original groupings (the abstract genres) and are applied ("assumed") selectively; (b) on the other hand, one and the same parameter can receive different forms when actualised ("fixed") through different mechanisms: These mechanisms concern the management of the semiolinguistic resources of a text.The correspondence between parameters and mechanisms is not fully reciprocal.In fact, the same parameter will be able to be brought up to date through different mechanisms, and it is in the specificity of the mechanisms that the singularity of each text is rooted.The textual output does not reduce itself to the mechanical application of an assembly of strict parameters.(Coutinho / Miranda 2009: 41) 1.3.Worthy of special attention in the context of the present discussion are the following formulations: (1) "foreseeable characteristics", also explicitated in the form "predictabilities" of genre; (2) a text "assumes" genre parameters, so that, as a result, the parameters are "fixed as empirical texts".
We are not told who is the agent that is supposed to foresee the defining parameters of a genre (is it the speaker?or the linguist?),or on what basis this predictive feat is/can be achieved.The underlying (tacit) assumption here is, I believe, that "abstract" genres are somehow "un-ontologically" extant prior to their realization, as a sum of parameters that can be predicted by virtue of some sort of (historical?/ a-historical?)regularity, and that "empirical" texts (what in Coserian terms we would call 'the text as ergon') have the ability to "assume" (take up, adopt) and "fix" some parameters, selecting them in different ways and putting them in different combinations, and also 'modifying' them ("bringing them up to date") through different mechanisms of realization.Again, we are not told if these "mechanisms" are "ontologically real" or not, and precisely where they are situated -in what kind of 'interstice' on the way between abstract genres and empirical texts.If, however, we check the list of examples of "mechanisms of textual realization", which are "the lexicon, the syntax, the enunciative organization, the mechanisms of material presentation such as the typography and the chromatic variation, etc." (p.42), it appears that these "mechanisms" are nothing else than the whole content of linguistic competence viewed on its various levels (idiomatic knowledge, elocutional knowledge, textual knowledge 11 ), with the crucial difference, however, that this knowledge and these operations are understood as nothing more than instruments for the "realization" in the textual products ("textual output") of an "abstract" pattern given in advance.
A serious problem one can notice with this account of genres is the lack of clear separation between what is considered to be the speaker's knowledge as such and what pertains to the description / explanation of this knowledge (with the different models that can be constructed about it).This leaves room for suspicion that the argumentation may contain surreptitious shifts back and forth between the two, and that the "epistemological and methodological" difficulties the proposed model is supposed to address are in fact engendered by the artificial projection into the speaker's knowledge of something that is exclusively a construct within one model of linguistic description of that knowledge.
For example, let us try to see exactly how the authors conceive the role of the speaker in the whole process of genre "realization".Aside from two explicit references, which will be examined presently, we can grasp it mostly as it surfaces from the overall view.Formulations such as the ones quoted above suggest that the speaker's 'intervention' 12 is reduced to combinatorial freedom -the freedom to choose some parameters over others from a preexisting "foreseeable" set, and the freedom to apply or not certain preexisting mechanisms in relation to various parameters.Thus, it can easily be noted that the underlying conceptual metaphor of this model is 'mechanistic' or, better said, 'robotic': the model indicates that combinatorial choices are being made, but the issue of why a certain choice is made, or what the reasons for a certain choice are, is not addressed within the framework of the model itself, but, rather, remains relegated to the endless variety of circumstantial factors pertaining to the "singularity" of each "empirical" text, or perhaps to the interplay of concrete pragmatic aims manifested in each act of communication via those texts.
Turning now to explicit references to this issue, the first suggestion that a model of the relation genre -text must account for "the active role of the subjects" is made by the authors in the context of justifying the advantages of their own model, by quoting P. Charaudeau's idea that speakers are able "to play strategically with the conditionings of genre" (Coutinho / Miranda 2009: 41).The second explicit mention of the "subject" is a statement that, by interpreting generic markers and thus identifying the "generic inscription of a text", "subjects-receivers" play the game proposed by the "subjects-producers" (p.49).It will be noted that both formulations remain within the range of the 'robotic' metaphor of effecting strategic choices in a game with pre-existing rules and "conditionings", and that a clear-cut separation between the speaker's intuitive knowledge (about genres) and the reflexive level of 11 For instance, if we try to translate this outlook into integralist terms, we can show that "the mechanisms of material presentation such as the typography and the chromatic variation" are non-linguistic components of textual expression (Texkonstitution) which only become relevant at the level of sense construction by virtue of "evocative relations" which support the correlation with certain units of sense in the given text.Thus, they are retrievable as components of the speakers' "expressive knowledge".12 For instance, the formulation that a text "assumes" some parameters is equivalent to tracing back from text-constitutional clues, in objective key, the fact that the text's producer "assumed" some parameters, which can also be recognized as such or "re-assumed", so to speak, by the text's interpreter.a description of genres (or exemplary models of genres) is still not drawn.As far as this issue is blurred, the matter of causally-oriented or purposiveness-oriented explanations (be it even for the combinatorial choices) will also appear blurred: neither clearly 'there', nor clearly 'not there'; somehow 'missing', yet somehow 'hanging in the air'.It is in this sense that I mentioned the 'ghost-like' nature of this problematics, and the fact that it simply will not 'go away', even when the investigations in question do not purport to tackle it directly.
1.4.It is obvious that postulating the existence of "abstract" genres as "un-ontological" entities is a theoretical and methodological decision that can only be understood as a sort of 'hypothesis' before the facts, and certainly not as something starting from the facts (the reality of discursive activity).The model is then constructed in order to solve difficulties and paradoxes generated by the hypothesis itself, and moves in circles, since a genuine solution is actually impossible unless the initial (unquestioned) hypothesis is disconfirmed and rejected.
If a framework like the one advanced by Coseriu is adopted, the point of departure for the investigative process is changed: instead of an artificial postulate of "abstract" genres, we would first direct our regard towards the "clear and certain" knowledge that is the speaker's linguistic competence, manifested in the activity of speaking and recognizable through its products.By so doing, the aforementioned difficulties will simply be precluded from the very beginning.
Regarded from the perspective of a realistic text linguistics, one that aims at explaining the speakers' actual knowledge for the construction of texts, text genres defined as "abstract" conglomerates of "foreseeable" generic features, devoid of "ontological reality", have nothing to do with the speaker's competence as an intuitive (technical) knowledge.The speaker is not really operating a selection out of a pre-existing "un-ontological" set 'floating' in a virtual reality somewhere outside his mind, but merely uses as raw material for the construction of his own discourse everything he finds at his disposal (in his own knowledge) and considers appropriate for the sense he intends to express -and this includes the evocation of one or more Discourse Traditions.For what can genuinely solve the dilemmas tackled by the authors of the study in question is nothing else than a concept of genre as (one type of) Discourse Tradition. 13 In the framework of integral text linguistics, text-typological knowledge (of which the knowledge of Discourse Traditions is one component) must be understood as part of expressive competence, and, to be sure, its degree of… "ontological reality" is precisely the same as that of the knowledge of idiomatic units and rules, as the «knowledge of things» and the «general principles of thought» that form the speaker's elocutional competence, as the knowledge of all the other components from the different sublevels of expressive competence.What appears to be a 'choice' out of several pre-existing options, or mere combinatorial variation, is nothing but a reified and simplified representation of the facts that: (a) linguistic competence is structured; (b) its components (units, devices, norms etc.) have functional roles proceeding from the specific principles of structuring on each level; (c) in discourse, the speaker makes creative use of each element, taking into account its functional status and making it work towards the realization of the sense he intends to construct.

13
For the concept of Discourse Traditions (DTs), their "secondary" type of historicity, and an in-depth analysis of both the relation between DTs and texts, and the change of DTs over time, see Kabatek 2005a and2005b.2. "Motivating" and "justifying" expressive knowledge 2.1.What does it mean, then, in concrete terms, for a linguistic investigation of (textual) sense to start from the speaker's own linguistic competence for the construction of texts (expressive competence as intuitive knowledge)?
The issue is far from trivial, because textual semantics is the one field most exposed to the danger of fading away into psycholinguistics, pragmatics or a theory of social institutions.Should we look at the question of what makes speakers construct sense the way they do, this tendency of collapsing textual semantics into other disciplines translates into a marked preference of researchers dealing with discourse/texts to skip textual-semantic considerations altogether 14 and jump to 'explanations' through non-linguistic causes and discourse-external constraints and circumstances.A telling example, as far as text typology is concerned, is J.-M.Adam's (1999) 'sociological' turn, if I may call it so: his claim that we should dispense with the notion of 'text type' altogether and recognize, instead, as operational at the highest (global) level of textual organization, only discourse "genres" defined as "types de pratiques sociodiscursives" (1999: 83, in a section suggestively entitled "En finir avec les «types de textes»").
Before, however, turning to such an extreme solution, it is worth taking a step back and examining whether a different view on the motivation of sense-construction orientations is not possible within the framework of a different kind of text-semantic description, one that starts not from hypotheses formulated in abstracto, but from the speaker's own expressive knowledge.

2.2.
With this aim in mind, let us examine the viewpoint advanced by Eugenio Coseriu in the monograph dedicated to linguistic competence (1988, section 3.3.8.), already summarized above, under 1.1.
The speakers themselves can and often do offer justifications of their linguistic knowledge as a technical knowledge.It is difficult, Coseriu emphasizes, to distinguish what is really part of technical knowledge as such, the knowledge of speakers as speakers, and what belongs to a totally different level, that of the reflexive knowledge of speakers as linguists (possibly only "amateur linguists", to be sure).Coseriu's suggestion as to where and how to draw the line of demarcation between the two is that only two types of "immediate objective motivation" are actually part of linguistic knowledge as a technical knowledge.These are the "historical motivation" (drawing upon the tradition of speech of one's community / communities) and the "functional motivation" (drawing upon the ways in which meaning and its expression are organized and correlated -upon the units, norms and principles of the functional organization of language at its various levels, which constitute the content of the speaker's competence): (1) The historical motivation: [...] kann der Sprecher auf die Frage "Warum sagen Sie so?" zur Antwort geben: "Weil man es so sagt."Bezieht er sich bei dieser Antwort auf eine Gemeinschaft und ihre gemeinschaftliche Tradition, d.h.er begründet sein Sprechen historisch und beruft sich darauf, daß er Mitglied einer Sprachgemeinschaft sei.(Coseriu 1988: 221-222) 14 Semantics is definitely present as idiomatic semantics -the lexical and grammatical meanings of idiomatic units, including the stratum of idiomatic procedures for text construction (textual "grammar") -, but not as textual semantics proper, i.e. the study of sense as the autonomous type of meaning specific of Level III in Coseriu's triad (see Table 1).
(2) The functional motivation: Bei der zweiten unmittelbaren Begründung gibt der Sprecher auf die Frage "Warum sagen Sie es nicht auf diese andere Weise?" die Antwort: "Weil dies etwas anderes bedeuten würde."Mit dieser Antwort bezieht sich der Sprecher auf die andere objektive Motivation der Sprache, nämlich auf die funktionelle Motivation.Er bringt nämlich zur Ausdruck, daß seine Ausdrucksweise gerade der Funktion entspricht, die er meint, und keiner anderen.(Coseriu 1988: 222) Should the speakers go one step further, i.e. try to "motivate the motivation", for instance by referring to one community in particular, in order to show exactly what that tradition is, or by indicating which particular functions, in what functional organization they have in mind, they will have entered the realm of reflexive knowledge, assuming the role of (amateur) linguists and, at the same time, the risk of error that this 'amateur' status brings along.It is, of course, the task of the linguist to identify the functions and their organization, as well as the extension of the various traditions of languages and texts, in order to give an objectively grounded justification to the speakers' technical knowledge. 15 We can now return to the initial question of this section (supra, 2.1.):Taking expressive competence (the speaker's intuitive knowledge for the construction of texts) as a point of departure in text linguistics means, first and foremost, that these two types of motivation which pertain to the speaker's own competence are the only "causes" and/or "purposes" we can deal with in the framework of a discipline of (text) linguistics aspiring at objectivity, i.e. at being faithful to its true object of study.By way of consequence, it also means that the task of attaching to them a properly grounded description and justification is the only type of "explanation" we can attempt to give.In other words, our realistic goal, in text linguistics, can only be to explore these two types of intrinsic motivation for the activity of creating sense.
Discourse-external conditions and factors of influence (pragmatic aims in any case, but perhaps also psychological and social factors), even those objectively attestable beyond the shadow of a doubt, will not actually justify how the sense of a text is articulated, although they may, of course, throw light upon the way a text is or could be used for interpersonal interaction in concrete circumstances.
I have emphasized in a previous paper,16 while discussing the issue of "external" text typologies, that contingent factors of a social, political, ideological nature etc., as well as ulterior motives of the individual speaker in a given practical situation cannot be taken as a legitimate object of study in text linguistics.For example, someone may create a poetic or a scientific text with the clear (discourse-external) practical goal of obtaining financial gainbut such a goal, even when attestable by objective evidence, remains outside the scope of text-linguistic investigation and will not represent a viable criterion of text categorization from the point of view of sense construction.To this someone may reply that, for example, in 15 As Kabatek (2003: 256) rightly points out, this is a distinction that concerns the finalities of the motivations / justifications given by speakers and linguists respectively, even though in practice the results may be different, and "mixed" cases can also exist: "[…] esta distinción de principios se refiere a las respectivas finalidades de las actividades y sólo indirectamente a los resultados.Un hablante «naif», sin ninguna formación lingüística, puede llegar a una clasificación lingüística adecuada que coincida con una clasificación científica objetiva.Y del mismo modo, un lingüista puede, a pesar de todo voluntad de coherencia y de objetividad, llegar a resultados equivocados.Pero la finalidad de ambas actividades es, en un principio, diferente, aunque empíricamente pueda haber mezclas: hablantes que apliquen una metodología y terminología practicamente científica, y lingüistas «naif»." the case of popular literature, if the publishing industry has a rule of paying authors by the number of words in a manuscript, an author wishing to gain more will write a longer text: in other words, that the external aim will actually shape, in a direct way, the linguistic constitution of the text.To this I will reply that in text linguistics we can never really start from the external goal, since this is not necessarily reflected in the textual constitution and does not directly correlate with any unit of sense: in the hypothetical case described above, the text in question will not represent, from our point of view, an example of a would-be 'class of texts written for financial gain', but a …long text pure and simple, and our task will be to explain how the sense is created and articulated in this 'long' text.

2.3.
On the level of discourse/text, the immediate historical motivation will manifest itself concretely as recourse to a tradition of text-construction, and can be most easily observed when the text is built with "repeated" units and rules, pertaining to text forms, categories and genres that are conventionalized or codified to a higher degree.The functional motivation involves only the identification of sense functions and their corresponding textual expression in the individual text, so the speaker may point out, for example, that a different way of expression would bring about the interpretation of a different meaning. 17 In a methodology strictly by the book, the two forms of immediate motivation have to be observed in the activity itself: in the way speakers construct their own discourse, in the tacit acceptance of adequate discourse, in the rejection, contestation or correction of inadequate discourse.However, since Coseriu's proposal implies a degree of distance or displacement between what can be observed in the activity itself and what kind of explicit motivations speakers would give about their discourse, I think it would not be devoid of interest to verify at the same time both aspects: (a) on the one hand, if the two types of "immediate motivation" can be validated in a real situation of discourse construction and interpretation; (b) on the other hand, what kind of differences appear between what the subjects actually say, and the way the subjects try to motivate the respective ways of expression.
The section that follows presents an attempt at observing and attesting these two sides of the issue of «motivation» in the domain of expressive competence.

An a experiment in text construction
3.1.In order to explore the speakers' motivation(s), we need to have several speakers construct a text with a sense that can somehow be objectively attested or even controlled beforehand: it is only then that we can observe their motivations in the activity itself (or, methodologically, in its product -the text they produced) and, at the same time, try to see how the speakers themselves motivate the way they said what they said.Ideally, if we have several speakers placed in the same situation and attempting to express the same sense, the differences in the texts they produce can be examined as clues to different motivations.17 Which type of immediate motivation will prevail in each case depends on the nature of the concrete textual component highlighted and on the typological characteristics of the text, given the peculiarity of the "loose" structuring of expressive knowledge on the whole (Coseriu 1988: 259).This "loose" nature holds true for the multifarious dimensions and norms of typological structuring of texts: "Das expressive Wissen ist [...] enorm verschiedenartig, und die entsprechenden Normen sind von ganz unterschiedlicher Verbindlichkeit.Sie reichen von den ganz allgemeinen Normen für die verschiedenen Arten des Sprechens [modalities of speech] über präzisere Normen für Textsorten [text categories] bis hin zu den sehr präzisen Normen für die Strukturierung bestimmter traditionell fixierter Textformen [text forms]."(Coseriu 1988: 159) How is it possible to control the sense beforehand, while also maintaining a situation that contains as little artificiality as possible (one that keeps as close as possible to real speech)?One answer may come from what is defined in integral linguistics as a process of speaking raised to the power of two, namely from translation: "a peculiar form of speech", "speaking by means of another language and with a content that is already given".In the case of translation, "the content to be expressed is given beforehand, up to its very details". 18 Translation requires reconstituting the original text in such a way that construction of the same overall textual designation and interpretation of the same sense become possible for the target-language reader.Thus, in fact, translation is not about conveying a 'disembodied' content, but involves conveying the text-constitutive procedures 19 and the sense of the original text in their interrelation.
For the purposes of the present discussion, I would like to bring an illustration by using some material that I collected as part of a wider project on the topic "Sense-constitutive sign functions and their relevance for a typology of poetic texts -a contrastive approach".In the framework of this project I devised a pilot qualitative questionnaire containing a total of 15 tasks (10 translation tasks and 5 reflection and self-analysis questions), which I administered on May 7, 2013 to a group of eleven 3 rd year students specializing in English at the University of Almería (see details below, 3.2.). 20The target group was chosen so as to have sufficient knowledge of the source language (English) in order to be able to fulfill the tasks of translation into the target language (in this case, Spanish), while remaining, however, as close as possible to the status of "naïve" speakers, i.e. not (yet) professionals in the field of linguistic or literary studies.
The fragment of the questionnaire that I intend to bring for exemplification here is focused on the following excerpt from Walt Whitman's poem Night on the Prairies (Whitman 1973: 452): I was thinking the day most splendid till I saw what the not-day exhibited, I was thinking this globe enough till there sprang out so noiseless around me myriads of other globes.

19
Textual constitution is made up of devices and procedures of expression that are specific to Text as Level III (the autonomous level of sense construction), and do not simply coincide with idiomatic units as such.For a detailed discussion, see Tămâianu-Morita 2012b and2013. 20 The project "Sense-constitutive sign functions and their relevance for a typology of poetic texts -a contrastive approach" was supported by a grant of Akita University within the «Researchers Overseas Transfer Program 2012», which gave me the opportunity to spend the spring semester 2013 at the University of Almería, Spain.All my gratitude goes to Jesús Martinez del Castillo for kindly accepting to host me for the duration of this project, and for giving me the chance to administer the questionnaire in one of his classes.I owe thanks to Toshiyuki Sadanobu (Kobe University), Marius Nagy (François Rabelais University of Tours) and Keita Ikarashi (Tsukuba University), without whose timely support I could not have gained access to numerous Japanese and French translated versions of Walt Whitman's poetry -some of them referred to in this paper, others to be used in further stages of the project.Last but not least: a word of commendation, as well as my expression of gratitude to the students who agreed to participate in the experiment and provided extremely detailed answers, with a wealth of material which calls for equally detailed analysis and interpretation.
(a) first, its status as a lexical innovation, identifiable as such at a pre-textual level (specifically, at the idiomatic level), by its relation to the systemic virtualities of the English language and to the norm(s) of English, both in Whitman's time, and for the contemporary reader; (b) second, its role as a unit of textual expression in this particular text, and its semantic contribution to the overall configuration of textual sense.
First, the created unit somewhat stretches the limits of the system itself, because in English "not" is specialized for syntactic negation, therefore typically combining with various verbal forms, as well as with adjectives and adverbs; by contrast, it does not readily combine with nouns to produce lexical negatives.Second, in the constitution of the poem, the pair "day" -"not-day" is placed in parallelism with the pair "this globe" -"other globes", a parallelism which not only enhances the idea of complementarity already present in the pair, but also induces a connotation of exhaustivity in time and in space respectively.After a lifetime of experience in the realm of the "day" (the visible, the mundane, the finite duration of individual human life), with the approaching moment of death, the poetic "I" awakens to the realm of "the not-day" (the hitherto invisible, the mysterious side of the universe hitherto hidden to perception, the cosmic infinity of time and space, death and immortality).Thus, in the poem, the term "night", with only one occurrence, in the opening line ("Night on the prairies"), designates the end of the "day", that is, the final moment on the axis of the first dimension (the visible and the mundane), the point in time where a window to the invisible world is open, whereas a new term had to be created in order to give an inkling of the second dimension -and this term is "the not-day".A faithful translation would have to create a textual unit in the target language that evokes the system and norm(s) of that language in the same way as the lexical innovation "the not-day" does with the system and norm(s) of English, and also fulfills an analogous function in the sense-articulation of the poem.
After a brief description of the content and organization of the questionnaire (infra, 3.2.), the results for the fragment in question will be presented (3.3.) and discussed (3.4.).We will thus see how the participants in the experiment attempted to express the same sense in Spanish, and what kinds of motivations they gave for their ways of expression.Section 3 will close with some supplementary remarks based on a cross-linguistic comparison of the proposed translation variants (3.5.).

3.2.
The pilot questionnaire was divided into two parts, administered in succession.In order to put the respondents in a situation where they had to make quick, spontaneous decisions, rather than dwell reflexively at length on each task, the time allowed for writing down the answers was 15 minutes.Some students needed an extension of around 5 minutes in order to finish filling out the whole questionnaire.
In Part I, students were asked to fill in blanks in a Spanish translation of several excerpts from poems by Walt Whitman.For practical reasons of time restriction, the full contexts (integral poems) could not be made available to the respondents, who were thus inevitably placed at a disadvantage in the process of understanding (=interpreting) the global sense of the texts.To compensate for this limitation, the fragments were identified by the author's name and by an explicit typological categorization as poetic texts, thus providing the minimal clues that would normally have been available in a genuine situation of reading and translation: "Part I.The following texts are fragments from the work of the American poet Walt Whitman (1819-1892).The questions below focus on certain lexical and grammatical meanings of the original, marked in bold.Fill in the blanks of the Spanish version with your own translation.
Note.You can also change other parts of the given translation, if you can think of a better variant."Thus, the parts left for fill-in were strictly limited to very short sequences, chosen for their potential difficulty from a contrastive point of view, 21 but students were allowed to modify freely any other part of the given Spanish version, in case they felt they could give a better version.
In Part II, respondents were asked to explicitate their understanding of the sense of the original and to motivate their own translation choices, by free answers written either in English or in Spanish: "Part II.Reflection questions.Write your opinion / impression freely, in English or in Spanish." The formulation "write your opinion / impression" was meant to suggest that they were not expected to make a 'philological' analysis or a fully justified 'literary interpretation'; instead, they were encouraged to analyze retrospectively their spontaneous and intuitive reaction to the texts.The explicit instruction was also given (and observed) that they were not to return to Part I or change their initial (spontaneous, intuitive) translation.
As already mentioned, in the context of the present discussion I will refer only to the task of translating the first line of the excerpt from Night on the prairies quoted above, with focus on the unit "the not-day". 22The task was given, in Part I of the questionnaire, in the form of a text with a corresponding blank: For this excerpt, the reflection question given in Part II was the following: "1.Regarding [B]: One published translation23 of "till I saw what the not-day exhibited" reads: "hasta el momento en que vi lo que mostraba la noche".Do you think this version is faithful to the structure and meaning of the original?Explain your reasons, thereby also justifying your own translation."21 Prior to selecting the units to be highlighted in the questionnaire, I examined an extensive body of 'candidates' by checking in detail different published translations into Japanese, Spanish, Romanian and French, and focusing on those textual units which seemed to pose considerable difficulties for professional translators, in most cases due to inherent differences of lexical and grammatical structuring in the target languages as compared to English.

3.3.
The proposed equivalents and the motivation given by the 11 respondents are presented in Table 2.The order of answers in the table is conventional, reflecting evident similarities in the strategy of translation adopted only for this particular fragment. No.

Proposed translation
Position regarding the equivalent "la noche" 1 el no-día I think "la noche" is a good translation because that's what he really means, but to be more faithful with the original one, in order not to loose that uncertainty that Whitman brings us without saying directly 'the night' it could be better translated as 'el no-día' or something like that.

lo que no es el día
A mi parecer, la noche podría ser una buena traducción, aunque es cierto que el autor no utiliza el término "night" que habría sido del todo explícito.Lo que ocurre es que el lector no es capaz de imaginar un espacio que no sea ni día ni noche, es decir, no podemos concebir algo que no sea ninguno de los dos.Asi es que, 'noche', al ser un perfecto contrario de 'día', para mí encaja.

aquello que no es el día
No, it implies something more than just a night ("till I saw what the night exhibited"); that is why I did not translate it as just 'night'.It implied all what the day is not: the 'no-día', but this structure is not appropriate for poetry, even for every-day language, so I wrote it as a periphrasis.I think the poem is not only about a day itself, but also about metaphysical questions.

lo que ya no era día
From my point of view it could be a possible translation but it does not show the real meaning or purpose of the author, who may refer to anything that does not form part of the day, not necessarily the night.

aquello que no era día
It is faithful in so far as 'the not-day' implies the contrary of 'day', which is 'night'.I tried to be faithful to the literal words shown in the text, that is why my translation would be 'aquello que no era día'.I think it is more specific, since we also have parts of the day as dawn, evening, in which there is no notion of night or day explicitly.

la noche
Yes, because the day has two parts: day and night, and what is not day has to be night.

la noche
Yo pienso que esa versión es correcta o concuerda con el significado del original porque creo que el escritor busca explicar o encontrar una palabra que signifiqe lo contrario al día, es decir, la noche.

night 24
Yes, I think it's quite faithful as it expresses the contrary thing of a day (because of the particle "not").

la oscuridad
This version is not faithful to the structure, but in accordance to the meaning.I think we can observe what the author wanted to say, even if in Spanish the structure is different.My own translation is more or less the same, but including more poetry and meaning.

el día ya consumido
In my opinion, 'la noche' does not give us a poetic sense in Spanish.
It would fit better if we avoid the easy way of translation and put another construction that shows that poetic sense.
Table 2: Questionnaire results regarding the translated equivalents for "the not-day"

3.4.
With all the inherent limitations of this very small-scale experiment, and difficult though it is to draw the line between the speakers' intuitive knowledge for the construction of texts (expressive competence) and their reflexive knowledge, it seems to me that the results presented in Table 2 do substantiate three types of observations.

3.4.1.
First, the content of the explanations given by the respondents in answer to the retrospective reflection question in Part II confirms that, at the level of the immediate motivation ("die unmittelbare Begründung"), Coseriu's categorization is both valid, and sufficiently encompassing.
(a) Statements like: "that's what he really means" and "in order not to loose that uncertainty that Whitman brings us without saying directly 'the night' it could be better translated as 'el no-día' or something like that" (no.1), "'noche', al ser un perfecto contrario de 'día', para mí encaja" (no.2), "it implies something more than just a night ("till I saw what the night exhibited"); that is why I did not translate it as just 'night'" (no.3), "creo que sí es una buena traducción al español, porque, en la versión original, se niega a día, es decir, 'not-day'; por lo que al negar esta palabra nosotros entendemos que es de 'noche'" (no.10) etc. fall into the category of functional motivations.
(b) Statements like: "this structure is not appropriate for poetry, even for every-day language" (no.3), "My own translation is more or less the same, but including more poetry and meaning."(no.9), "'la noche' does not give us a poetic sense in Spanish" (no.11) fall into the category of historical motivations, more precisely recourse to the knowledge of a tradition of texts -here, the tradition of 'poetic discourse' in its widest sense.
These were the only types of motivations offered by the speakers themselves, and it is obvious that these indicate with precision the main concerns that were at play when the students interpreted the English original and, subsequently, when they constructed their equivalent texts in Spanish: to convey the 'right' meaning with 'the right words', and to do it in a fashion that is in accordance with the tradition of poetic discourse.Answer no. 3 also shows an interesting example of a situation where these two motivations are in conflict, and the subject gives priority to one of them, sacrificing the other: the respondent considers that functionally something like "no-día" would have been necessary in order to express everything that "the day is not", but that "no-día" "is not appropriate for poetry" or even for everyday (non-poetic) language use, and therefore decides in favour of a periphrastic construction, "aquello que no es el día".

3.4.2.
Second, we can notice the degree of coincidence or, on the contrary, the distance between what speakers actually do (their proposed equivalents, i.e. the knowledge manifested in their activity of speaking and visible in the text they created) and what the speakers think they do, especially when their explanations step into the domain of reflexive knowledge.
I find the immediate (functional) motivation given for version no.625 extremely interesting because it is a direct manifestation of the speaker's (idiomatic) knowledge of the way the pair "day" -"night" is semantically organized in English, with "day" as the neutral term (24h-duration) inclusive of "day" = <time of daylight> and "night" (see, in this sense, Coseriu 1973Coseriu /1981: 233-234): 233-234).The same can be said, in fact, of explanation no.5: "we also have parts of the day as dawn, evening, in which there is no notion of night or day explicitly" (emphasis mine -E.T.-M.).In these two cases, what the subjects have in mind is the idiomatic organization of significata and how these significata are (/can be) used for designation, i.e. the "meaning" at a pre-textual level.By contrast, the motivations given for no. 4 and no. 3, also of the functional type, regard the level of "meaning" at the textual level, that is, the sense created with the help of this unit in this particular text, suggesting the second dimension of temporality, different from that of mundane experience and human time-scale: "the author […] may refer to anything that does not form part of the day, not necessarily the night"; "the poem is not only about a day itself, but also about metaphysical questions" (emphasis mine -E.T.-M.).There is no contradiction, therefore, between no. 3 and 4 on the one hand, 5 and 6 on the other.
These instances illustrate precisely the fact that, when available, the immediate motivation given by speakers, in addition to and in conjunction with the actual discourse they create, can provide clues for an objective justification of sense-construction processes in the framework of text linguistics.
At the same time, it is worth noting that, when asked to evaluate a published translation, the answers of the respondents who gave a different version themselves are often formulated in a diplomatically oblique fashion: "I think 'la noche' is a good translation because that's what he really means, but…" (no.1), "A mi parecer, la noche podría ser una buena traducción, aunque es cierto que el autor no utiliza el término "night"..." (no.2) , "From my point of view it could be a possible translation but it does not show the real meaning…" (no.4), "Desde mi punto de vista creo que sí es una buena traducción al español […] [BUT] Mi traducción ha sido: not-day=oscuro día…" (no.10).Even when the respondent's own proposal is radically different (and in fact better) in terms of sense-construction strategy, the comment begins with a declaration of compatibility: "My own translation [= "la oscuridad"] is more or less the same, but including more poetry and meaning" (no.9; emphasis mine -E.T.-M.).These are all instances where, clearly, the subjects' actual manner of expression in the discourse they created (and the motivation following the "BUT") are the genuine and reliable starting point for a justified explanation, in the framework of text linguistics as a linguistics of sense, of the sense processes in question.

3.4.3.
Finally, the questionnaire results allow us to point out in a concrete way the difference between discourse-internal motivations and discourse-external conditions, reasons and purposes.Admittedly, the students who agreed to take part in the experiment may have (also) had their own private reasons, ulterior motives and practical limitations for giving the translations they did and formulating the justifications they did.For example, some respondents may have wished to impress the teacher or the questionnaire proponent by clever and original answers, possibly in the hope this might indirectly contribute to getting a higher grade, 26 or, in reverse, some may have had to expedite the answers due to the physical lack of time (thus giving a variant more or less at random, or one that was not truly a response to the request, such as no.8 in Table 2.).Are such motives relevant for a grounded text-linguistic explanation of the variants provided by the respondents?The answer is: NO, they are not.
Let me clarify this point by referring to what I find to be the most obvious example: In the framework of text linguistics as a linguistics of sense, the genuinely objective explanation of no. 8 simply cannot be relegated to an external condition like physical lack of time, to something psychological like a lapse of attention, or to something pragmatic like the aim of challenging the questionnaire proponent. 27Neither can it be dismissed as an error pure and simple.I believe we can safely describe the text-semantic mechanism at work here in the following terms: The respondent treated the given request not as an exercise in text-construction proper (i.e. the construction of a Spanish text matching the sense of the English original), but as a kind of disambiguation game, a "puzzle" to be solved by providing a possible designation 28 of the innovative lexical unit the not-day, a designation which was then indicated (pointed at, labled) by way of the English word night as such.This is as far as text linguistics proper can and will go, inasmuch as an explanation of this sample text (variant no.8) is concerned.The speaker him/herself (the respondent) provided an "immediate motivation" of a "functional" nature: assuming what (s)he believed to be the quest of the author of the original, the respondent too was looking for "the contrary thing of a day".There are, however, many different levels or components which can be addressed by or involved in such an immediate "functional" motivation: thus, what the text-linguist can explain is the underlying mechanism of this little discourse, with the levels involved (the shift from the level of sense back to -or down to -the level of designation) and its purpose (designational disambiguation).

26
Naturally, the respondents were informed that the questionnaire would have no bearing whatsoever on their academic grades.This does not, however, preclude the formation of at least 'unconscious' hopes of this nature.

27
The list of conceivable psychological reasons and pragmatic motives is endless: the respondent may have disregarded the translation request on purpose, in order to defy the questionnaire proponent; the respondent may have intended to voice a protest, for instance because (s)he considered the question too difficult; a combination of various conscious and unconscious reasons might have been at play, and so on and so on.It is doubtful whether even a full-fledged psychological approach could actually unravel the 'true' personal reasons that led this particular individual to act in this way in this particular situation.But even if it were possible to pinpoint the psychological reason behind this discourse, this reason does not, in itself, provide any explanation of how the discourse is constructed internally, and therefore cannot answer the question of why it makes the sense it does.

28
-or, to be exact, what the subject in question obviously considered to be the designation of "not-day" in the given poetic text.

3.5.
Interestingly enough, the different versions proposed intra-idiomatically by the Spanish students correspond, in terms of text-constitutive strategies, with various options of translation viewed cross-idiomatically.This is not surprising, given that the task was not to simply create a lexical equivalent in the abstract, but to translate a text -i.e. to convey the sense for which the respective lexical unit is only an instrument (a means) of expression.
In Léon Bazalgette's French version, "le non-jour" 29 corresponds, in terms of technique of sense construction, to variant no. 1 in our questionnaire results.So would, without a doubt, a Romanian translation as "ne-ziua", 30 but applying this procedure would automatically entail, for the present-day reader, an additional intertextual evocation (one that does not exist in the English original). 31By contrast, Francisco Alexander's Spanish version proposes "la noche", like the students' variants no.6 and 7. 32  As for the Japanese translation, this apparently simple lexical creation must have raised almost insurmountable difficulties.Although the Japanese system offers a wide variety of productive negative prefixes, with partly overlapping morphological regimens, such as hi-, hu-/bu-, mu-/bu-, mi-or han-33 , none of these can be easily combined with the first equivalent for "day" that comes to mind, namely the very short "hi" (the core signification of which is <sun> and <sunshine>, and thus, by extension, <day>).Moreover, the nouns which conceptualize the designational area of <day -night> have a very different organization, and there is no single unit that could cover all the variants-in-use (Sprechbedeutungen) of the English "day" (neutral term in the opposition "day" / "night", a point on the axis of time, interval of time, period of the presence of light, opposed to "night" as the period of darkness).Of the two published translations that include the poem in question (Whitman 1971(Whitman /1981(Whitman and 1959(Whitman /1999) the former, by Norihiro Nabeshima & Masayuki Sakamoto, resorts to the periphrasis "hiru denai mono" (approx.«what is not day [lit.'noon'=daytime]»), 34 like versions 2, 3, 4 (with a slight difference) and 5 in our questionnaire results, whereas the latter, by Shigetaka Naganuma, simply makes use of the antonym "yoru" (approx.«evening/night») instead. 3529 "Je croyais le jour le plus splendide jusqu'au moment où j'ai vu ce que le non-jour montrait, / Je croyais ce globe suffisant jusqu'au moment où, en un tel silence, jaillirent autour de moi des myriades d'autres globes."(Whitman 1955: 220).

30
The published Romanian translation by Mihnea Gheorghiu (Whitman 1956) does not comprise the poem in question, but I think "ne-ziua" would have been the logical choice, given the high productivity and versatility of the negative prefix ne-in Romanian (for a synthetic presentation with numerous examples of usage, see Zafiu 2004).

31
While the derivation with ne-in the case of nouns did exist in the norm of the language (ex."neom"), albeit with restrictions and mostly for uses of characterization, or predication of a property, the procedure was reactivated and reinterpreted by the 20th century poet Nichita Stănescu as a general possibility extended to any noun and with all the syntactic restrictions suspended, through his 1969 volume entitled Necuvintele ("The Not-Words"), which enjoyed enormous popularity, as well as dissemination among a wide range of readers, through compulsory education textbooks and cultural programs in the media.
4. Signifying finality and speakers' "intentions" vs. discourse-external goals 4.1.When dealing with investigations on speech/discourse, which, given the multifaceted nature of this object of study, often involve interference or partial overlap of frameworks from diverse sciences, distinguishing «causal» (and deterministic) explanations from «final» (and non-deterministic) explanations is, of course, necessary, but not sufficient.For instance, while physiological (including neuro-chemical) factors and psychological factors (perceptions, affects, emotional states etc.) may act deterministically as efficient causes at the level they pertain to, i.e. the biological 'infrastructure' that makes up the individual human being and therefore inevitably constitutes the backdrop of cultural linguistic competence and of the activity of speaking, 36 speech/discourse-external goals and pragmatic aims surely belong to the realm of finality and are (or should be) recognized as manifestations of the speaker's activity in the realm of freedom, not in the realm of natural necessity. 37Therefore, in order to delineate what kind of explanations belong to the proper scope and framework of text linguistics, one more line of distinction has to be drawn.
No one can deny the objective reality of psychological and social factors and conditions that influence individual discourse, just as no one can deny that texts are sometimes used purely instrumentally, for transmission of factual information, or as the more expedient substitute of an action.The point that needs to be emphasized is that, real and objective as such factors may be, they will always remain outside the scope of a linguistic explanation of discourse/texts.In text linguistics, the question "Why do you say so?" can only be translated as: "How do you say it in such a way that your signifying intentions are conveyed to the interlocutor?",and correlatively, "How does it become possible for the interlocutor to make sense of the text?" 4.2.In this context, it is worth reexamining the sense in which terms like "intention" and "attitude" are used in the framework of E. Coseriu's integral linguistics.The "intentional character of language" is directly observable, in the first place, de manera inmediata, simplemente en el sentido de que se puede hablar o callar y de que se puede hablar de distinto modo, según la intención.(Coseriu 1967: 36; emphasis mine-E.T.-M.) 36 For the importance of separating cultural linguistic competence from the biological faculty of speech, see Coseriu 1955-56: 287 and1988: 69. 37 Sociologists might be able to persuade us that, in very peculiar situations, individual speakers are also subjected to 'forces' from the outside (social pressure, censorship etc.) that appear to act as determining causes, in addition to 'forces' from the inside, of a psychological nature, or perhaps by way of extreme psychological pressure.Even in such cases, however, we are not dealing with anything of the kind of natural causality: the way speakers speak still remains in the realm of freely assumed decisions, albeit in circumstances where a perverted, destructive type of "intersubjective" norm is at play.The finality (purposiveness) that belongs to the essence of language is the signifying finality, by virtue of which language is defined as logos semantikos, as creation of meaning. 38 However, pure semanticity is never given as such: it is always determined as creation of signification, designation and sense.In turn, at the level of discourse (Level III in Coseriu's triadic model), the construction of sense is a manifestation of the speaker's "attitudes": [...] todo discurso, todo lo que se dice, tiene no sólo designación y significado, tiene también sentido.Se dice algo, para manifestar una actitud, para establecer relaciones particulares con el interlocutor, etc. (Coseriu 1987a: 22-23) By virtue of these "attitudes", in discourse acts, logos semantikos is further determined (= is orientated in its actualisation) through discourse-specific finalities of diverse ranks.I have argued in previous studies that it is these discourse-specific finalities that constitute the basis of text-typological organization, vectors that delineate the 'typicalness' of genuine texts at various depths. 39The most basic (primary or first-degree) discourse orientations can be conceived in the form developed by Coseriu (1948Coseriu ( : 3, 1957Coseriu ( : 247, 1987b: 11) : 11) on the basis of the Aristotelian triad of the "apophantic", the "pragmatic" and the "poetic": (1) the scientific or ("logical") finality -the determination of discourse as logos apophantikos through the values of truth and falseness; (2) the "practical" finality -the determination of discourse as logos pragmatikos through the values of practical effectiveness; (3) the "poetic" finality -maximal manifestation of the signifying function itself, without restrictions of truth or efficiency.
The apophantic, poetic and practical orientations reflect the speaker's "intention", or "attitude" in given discourse acts ("el propósito del sujeto lingüístico en un acto determinado"), 40 and must be understood as inseparable from textual sense.A nuclear example of this basic triad of finalities and the effect they have on the construction of sense is given by Coseriu using the utterance Socrates is mortal: at the level of discourse / text (Level III), three distinct paths of sense construction can be unfolded, according to the text's finality, that is, if its overall function is: (i) the conclusion of a syllogism, (ii) a warning addressed to Xantippe or (iii) a small poem (cf.Coseriu 1957: 247).
It is thus obvious that terms like "intention", "attitude", "purpose" or "relation with the interlocutor" do not concern psychological states or pragmatic aims and relations, but intentions and relations of the speaking subject exclusively as a subject with a view to his/her own speech.4.3.These orientations (signifying "intentions" or "attitudes" of the subject) are inseparable from the process of discourse construction and from the sense thereby created, expressed and understood.They give the impulse and establish the inherent direction for the unfolding of the process of sense-construction, and can therefore properly be called "discourse-internal" orientations.Sub-categorizations are, of course, possible, and remain one of the most pressing tasks of integral text linguistics. 41 By contrast, causes or triggers derived from psychological factors, as well as external aims, purposes or goals set by the individual speaker in the domain of practical action 42 areat best -of secondary importance: the former can vary indefinitely without leaving any trace whatsoever in the constitution of the text; the latter belong to the conditions of use of the text, and are therefore disjunct from discourse-internal finalities.
It is in this sense that we must understand, for instance, Coseriu statement (1987b: 17-18) that "political texts" cannot represent an autonomous class for text linguistics as a linguistics of sense, because they are separated as such only by virtue of their extra-linguistic content and contexts of use.From this standpoint, external classifications of texts according to the social situations or institutional contexts in which they are used cannot represent, strictly speaking, a textual typology, and cannot be constructed in the framework of text linguistics, precisely because such a typology merely comes down to a classification of social situations or institutions, which other disciplines are more qualified to operate.

4.4.
Last but not least, let me address one point which, if taken at face value or out of context, may seem to etiolate the consistently finalistic position that Coseriu maintains throughout his work.
In the explanation about the two processes that make up linguistic change, namely "innovation" and "adoption", the former, which is a phenomenon specific of particular acts of discourse, is recognized to occur not only for a certain finality, but also, possibly, owing to actual "causes" (determining causes of a psychological or physical nature): 43 41 An excellent example is Mircea Borcilă's subcategorization of the poetic modality of sense construction (Borcilă 1981(Borcilă , 1987)).For a brief presentation with English examples, see Tămâianu-Morita 2012a: 17-19.42 Coseriu (1957: 251, footnote) insists upon the need to avoid confusion between this domain and the essential finality of language: "[…] no debe confundirse la eventual finalidad exterior del lenguaje, su instrumentalidad, con la finalidad significativa que pertenece a su esencia".

43
The arguments quoted here from the 1983 study devoted to the problem of linguistic change, published in English, had already been made, in essentially the same form, in Coseriu 1958Coseriu /1978: 80. : 80.As a general principle, innovation as such can also be unintentional, that is, it can also be «causally» conditioned in a real sense, for speech is also a psycho-physically conditioned activity.(Coseriu 1983b: 58) When such causes are at play, then the innovation must be understood as "a deviation from freedom due to physical necessity". 44Nevertheless, "adoption" can only be motivated finalistically: Adoption, on the contrary, is an exclusively mental act that takes place in the language as a technique, that is, on the level of «linguistic knowledge»; therefore it is always intentional though intuitive (that is in principle no different from usual language learning), and so it can have no «cause», but only a final (functional, cultural, social or aesthetic) motivation.(Coseriu 1983b: 58) The role of adoption, with its motivation through finality, is always essential as far as linguistic change is concerned, and that is why "a causally conditioned innovation surely has little chance of being adopted and spread: there are for instance no examples of slips of the tongue that would have been generalized" (Coseriu 1983b: 58).
Thus, as far as the continuous systematization and re-systematization of a language (Level II) is concerned, it is clear that final motivations always remain in the leading position.What about the level of discourse itself (Level III)?How are we to understand the relative weight of «causal» conditioning factors vs. «final» motivations in the creation, expression and interpretation of sense?
I think the answer is obvious from the very nature of textual facts generated by the causal effect of a psycho-physiological condition: such textual facts are accidental -they are, in a strict sense, exceptions to ("deviations from") the rule of "freedom" and purposiveness.The fact that they must be understood as subordinated to the principle of finally-driven expressive activity is backed up by the existence, at the level of textual competence, of a whole array of sense-construction procedures to be used precisely in order to deal with such instances: if the speaker himself recognizes the accidental error or unintentional expression, he will amend it by strategies of self-correction; in dialogue, the hearer may ask for clarification and/or modification; sometimes such instances are the source of unintentional humour, perceived in this way by the interlocutor etc.On the other hand, it is also possible to take up psychologically-determined processes and elevate them to the status of sense-construction principles, in certain text types and text-construction traditions.The technique of écriture automatique in surrealism is one example in point.In this case, the product of the causal process is merely a raw material for textual constitution, and is subordinated to the text's internal finality. 45 However, when the causal factors gain the upper hand and actually foreclose intentionality, we enter the realm of abnormal speech / communication, the realm of lack-of-sense (breakdown of sense-construction), and therefore of non-text. 4644 "como desviación de la libertad debida a la necesidad física" (Coseriu 1958(Coseriu /1978: 80: 80; translation mine -E.T.-M.).

45
Using the terms of Mircea Borcilă's sub-typology of poetic texts (1987), this can be described as a particular case of the "asemantic-asyntactic" principle of sense-construction.

46
The linguistic output connected with the various psychopathological conditions is precisely that: "output", a linguistic product directly caused by factors pertaining to natural necessity, an "ergon" without "energeia" and "dynamis".Its proper study does not lie within the scope of text linguistics, but only within the powers

Post-scriptum: Reviving the (human) subject
Looking back towards the recently concluded 20 th century, with its mainstream linguistic trends ultimately falling prey to various kinds of positivistic reductions, we can only agree with Takashi Kamei who, in his explanatory afterword to the Japanese translation of Sincronía, diacronía e historia (1981), rounds up his overall characterization of 20 th century linguistics with the bitter exclamation "Linguistica in absentia hominis!". 47If, as Johannes Kabatek (2012: 104-108) suggests, there is hope that the 21 st century may bring along a clearer understanding of the possibilities and limitations of the new empirical trends, which benefit of technological advances of a magnitude never known before, then the time also is ripe, perhaps, in counterpoise, to rehabilitate the speaking subject as the true protagonist and focal point of all linguistic inquiry, where "linguistics" proper equals «linguistics as a science of culture».
For this field -our field -, the 'speaker' does not coincide with a collection of neuronal connections, no matter how real or how complex these may be.The 'speaker' we are concerned with is neither a cyborg, remote-controlled through the most sophisticated neuro-chemical command system in existence in the natural world, nor a slave at the mercy of all-powerful social forces.The time is ripe, in other words, to revive the speaker as a human being endowed with reason (= rationality), self-awareness and intentionality -all, paired with a constitutive dimension of intersubjectivity from which freely assumed expressive commitments derive and are put into action every time the speaker speaks. 48 It is with this human subject and his creative activity (creation of significations, designations and sense) that we deal in the realm of linguistics proper: anything below or beyond that will simply constitute a totally different object of study, and thus require the attention of totally different scientific disciplines.

"
[B]  (Night on the Prairies) I was thinking the day most splendid till I saw what the not-day exhibited, Yo pensaba que el día era lo más espléndido hasta el momento en que vi lo que mostraba _______________ ,"