TIME MODELS IN ESTONIAN TRADITIONAL AND LITERARY BALLADS


Arne Merilai
1. From synthesis to analysis
There was a time when I kept thinking that ballad was the base of the European mind. European life seemed to be like a ballad to me, and I think I was not very wrong in that. Now I am studying the pragmatic linguistic analysis of poetry (deixis, speech acts, implicatures) and I have given up the synthetical method. But thanks to Mr Tom Cheesman who asked me to restore that forgotten intercourse. I became interested again, for the lapse of time presented a new viewpoint for me.
The current paper was presented at an international ballad conference, held in Wales on July 18-24, 1996.

1 On these points see in particular Keith Green (1992), John R. Searle & Daniel Vanderveken (1985) and naturally H. Paul Grice (1989).
My early images about ballad as a synthetical form are summarized in the following quotation:
It is impossible to imagine the development of Estonian poetry without lyroepics or ballad.
/---/
The concept of ballad contains many different phenomena: Old French ballad, lyroepic folk song and romance, broadside ballad and Bänkelsang, lyroepic or even lyric artistic ballad. The genre also exists in music and film, in journalism and on the stage. In addition to the enormous bulk of works in this genre, the ballad is very widely spread - both historically and geographically. All this indicates that we have not to do with a neutral phenomenon in history.
Ballad emerged as a Provençal dance song. Its origins are in folklore, ecclesiastical art and classical antiquity. Along with the spread of ecclesiastical lyrical poetry the Old French poetry had an essential effect on the fading of European local epic (syncretic) tradition. The lyroepic ballad is a neosyncretic genre that consciously synthesizes lyrical poetry, epics and dramatics, which had meanwhile be come autonomous. /---/ The substantial ground for its genesis lies in the spread of modern consciousness and in the emergence of the subject from the society: The separated lyric and epic germs in their antagonism form the source of the ballad. The influence of the ballad's structure and poetics on the later literature is significant. From the 15th to the 17th century the ballad was one of the main forms of people's aesthetic consciousness in Europe, i.e. it was universal. The popular newspaper journalism was established under the influence of the broadside ballad's boom; the 18th-century Bänkelsang was an important predecessor of the film. The budding romanticism, relying strongly on the popular ballad, created the literary ballad. The genre is most actively practised just before the breakthrough of romanticism and its manifests /---/. In the ballad there is much revealing for the later poems, lyric novels and fate-dramas, for the romantic tradition in general. There is a dramatic conflict concealed in the epic contents, expressed in a lyric form and in the interests of the lyrical. The spirit of the ballad passes through literature just as the spirit of romance passes through romanticism.
The literary ballad is a lyroepic genre that is characterized by a closed synthetic structure where the lyric (subjective), epic (objective) and dramatic (conflicting) ideas have different functions forming a densely interwoven whole (according to J. W. Goethe Urphänomen - Ur - Ei). /---/ Its totality of arts (according to G. W. F. Hegel) points to its epoch-creating role for the modern system of arts, just as epic created the antiquity (Merilai 1991:136-137).

Although John R. Searle has noted that Hegel, the main bearer of balladic mind, is "a dead horse« on the contemporary market of philosophy, the ballad has no possibility to give up the Hegelian syntheticism. The object is still synthetic, but there is no need for the synthetic research of ballad. The researchers have be come more analytical in their work. The more they will use logic as the method of re search, the better the result will be. I am sure there will be a time when the research of ballad will transfer to the line of Frege and Russell. The incredibly large amounts of material collected demand that the synthetic approach be replaced by more general and rigorous approaches of formal logic. This will involve the deconstruction of the prevalent paradigm, of course, but who would regret it?

2. Ballad presents time
Although my presentation is perhaps not as formalistic as usual, it justifies the ideology of formalism. For a comparison I would like to quote my conceptions of the earlier period. At the same time I call on to negotiate them.
Now that I have got used to the analytical viewpoint, I have a new comprehension of the essence of ballad as an object. I believe it is common sense. What is a ballad? Any ballad is a narrative. It tells us about an event, maybe a dramatic one. It claims that things in some situation are so-and-so, that they have certain preconditions and corresponding consequences. Whatever the story is about, its defining metaphysical theme is time. The ballad expresses time by presenting an event with its prologue and epilogue. But what kind of time?
Presumably it does not present the mythological circulating time. It is well-known that the ballad rebuilds mythological songs, deleting all that is typologically unsuit able. Obviously the ballad does not present the feeling of eternity either, which is characteristic of religions. Legends are closer to religion, but they are apt to leave the concept of ballad. Probably the ballad presents neither the modern relative nor the multidimensional time. The ballad rather presents the time which remains between the times noted above - it is the linear feeling of time of modern (or modernizing) man. It certainly is the most widely spread system of time - past-present-future (H-P-F). That is, the linear time which moves from the past to the future through the present. Telling a story, be it either about external or internal reality, the ballad explicates the linear moving of time. Telling a story, the ballad presents the time on which the story has been "pasted". The ballad is the story pasted on linear time.
That the ballad presents the modern linear time can principally refer to the circumstance how European ballad made its way into the very old animistic Estonian folk lore in the late Middle Ages, which, by the way, already clearly formed in the times of Homer. We, Estonians, are European Red Indians, the aborigines of this continent who have resided here for more than 9000 years. Of course, the ballad met some resistance by the mythological and parallelistic time conception, trying to assimilate its elements, rename them or leave them out at all. One can hypothetically show how linear time adjusts itself in Estonian folklore, how it becomes prevalent there and how it moves on naturally into Estonian literature. It can be claimed that the concept of historical time, alien to Estonian language, spread in Estonian literature of the 19th and the 20th centuries, partially thanks to the pursuit of the romantic fate ballad (Fr. R. Kreutzwald, J. Bergmann, J. Tamm, M. Under, A. Kallas). At the same time some trends of development of Estonian poetry leave the balladic time, breaking themselves loose from the glue of linear time. This is, first of all, the case of A. Alliksaar's poetry.
So ballad as presenter of time is not universal; it is connected with a certain conception of time. Although ballad is not universal, the balladic system of time is maximally dominant in Estonian literature. However, in the boundaries of that common conception, there can appear different historical models. I have been trying to take note of them, but a more exact research can bring about some additions.
Thus the ballad presents the common W. European concept of time. My first idea was to investigate some time models in the classical ballads by Child, but the organizer of this conference, Mr Cheesman asked me to concentrate on Estonian material, because it is less known to the audience and therefore might be more interesting. I appreciate the advice, for I know the Estonian language and Estonian material better. At the same time Estonian material is generally quite similar to the European one because of its balladic time conception. As to the stipulative models to be discussed below, I would like to quote Tom Cheesman's letter as expert opinion:
Many thanks for your letter and abstract. Your paper promises to be extremely interesting. I anticipate some difficulties of comprehension, since you are A) referring to unfamiliar texts and B) applying an unfamiliar mode of analysis, as far as most of the audience is concerned. But for both reasons, the paper will be an excellent, thought-provoking contribution on relations between traditional and literary ballads.

He also adds a thought which has an especially important message for me as an Estonian scholar. The thought might provoke pros and cons among specialists of other countries too:
It certainly seems to me that ballads, as a form of narrative, normally require a clear temporal system, and I also suspect your systems I - IV are all found in "traditional" / folk ballads, at least in W. European traditions (Cheesman 1995).


3. Time and narratology
Analyzing Estonian traditional and literary ballad it becomes evident that both of them present the linear running of time. But it can take place in historically different forms. I have distinguished five basic types of expression, but a more detailed classification might be possible. I would like to emphasize that I consider the appearance of time as the metaphysical base of the narration and overlook the length of ballads, the number of stanzas or whether the number of scenes is large or small, as well as the aspects belonging to the physical side of the story. The latter, although it influences the time for the presentation of the ballad, does not mean that the longer story and the shorter story could not represent one and the same metaphysical model of time.
Neither do I mean the span of the time period depicted in the content of the story. The content of the story can include longer or shorter segments of time (it can concentrate on the whole life story of the hero or select only a very small part of it, etc.). The action can take place in one or in several days and so on. Treating the time as such I do not observe the narratological problems of depicting the time. The plot of the story can contain pro- or analepsises (looking forward or backward), it can be elliptic or resuming (accelerative), scenic or descriptive in style (one-to-one or retarding). The presentation of the story can be singulative, repetitive or iterative (single, repeating or concentrating presentation of an event). I do not mean the interesting problem of poetics called leaping and lingering, but the time in a very general sense of the word. What I mean is that the ballad, whatever its narratological structure, always presents the principal moving of time from the past to the future. Any ballad presents a story with a (presupposed) reference to the prologue and to the epilogue, presenting the situation in present with its causal connection with the past and the future. It presents either the linear or the vectorial moving of time.
In some possible ballads it could also happen that the protagonist transfers in time and appears to be his or her own father or mother or vice versa. In that case the development of his personal story has simply stumbled on the progress of time, generally accepted. The balladic make-believe does not produce such a wild fiction, of course, but tries to suggest truthful stories, and therefore it does not experiment with the time of the plot. Ballad is doubtlessly an admirable object of the theory of fiction (for which see Currie 1990).
Let us observe the following time models by some examples from Estonian traditional and literary ballads.
These models are not rigid, it is always possible to replace them by some more precise time or modal logic system, e.g. like that presented by Graeme Forbes (1985).

4. Linear time system
 fig1
The peculiarity of this type is that the presentation of an event may sometimes in cline toward parallelism. One and the same event can be presented again in a slightly changed form as an alternative event without any overt development of time. In our days one is liable to consider this case simply as a repetitive, but then the ancient cognition of time, to which the parallelism seems to be pointing, would get lost. Obviously the parallelism suggests that the following different moments of an event can be perceived as if existing parallel in time (P1 = P2 = .... = Pn and probably not P1 < P2 < .... < Pn). Still, in spite of the ancient tendency to parallelism, the main stream of Estonian ancient trochaic-metric ballads is linear, i.e., moving from the past to the future, with a more thorough observation of the present. This linearity could not have been the time background of ancient Estonian folklore of the pre-balladic time. Probably it began to spread in Estonia through Scandinavian and Germanic contacts. My hypothesis is that Estonians had to acquire the linear time model via other, foreign, patterns.
In Estonian traditional poetry parallelism is the basic poetical principle on the level of detail, motif or scene. More hypothetical is the parallelism reaching the expression of time. Not every researcher may agree with this thought. In that case the corresponding stories can be rather classified as belonging to Model III. I myself am apt to think that the repeated haggle scenes between the bridegroom and the dead in the ballad Grave's Girl `Kalmuneiu' express rather psychical simultaneity than following one another in time. Neither does the concept of gradation always include the idea of the sequence in time - it can express simultaneity as well. The most illustrative ex ample is Husbandslayer `Mehehukkaja' in which the episode of the flight of the woman into the forest and her repeated requests for help from the swine, the well, the birch and the asp refer not so much to the sequence in time of the events as to simultaneity. The same applies to the ballads Skydaughter `Ilmatütar', Bean and Pea `Uba ja hernes', and especially to Ransomed Girl `Lunastatav neiu' where the girl, violently taken to a Russian ship, dramatically haggles over herself with the kidnappers. The audience has to go through the embarrassing situation several times, with only small changes. Of course, for the oral tradition parallelism is an effective mnemotechnical means, but it is also rather connected with the corresponding time concept than excluding it. At the same time it is plausible to suppose that the noted ballads just express the conflict between the two historical notions of time in the period when the parallelistic model is regressing and the linear model is progressing. For example, the ballad Mareta's Child `Mareta laps' appoints to Jesus in its deus ex machina solution. In Estonian conditions the fact proves the later origin of the ballad. The same ballad is notably more time successive, that is, more linear in time by its parallelism. (For Estonian folk songs with translations see Kurrik (1985).)
A contemporary reader can naturally dispute about it. Parallelism as an archaic feature undoubtedly hindered the fast linear motion of time, but the latter as a supposed feature of the free market was in fact hindered in real life in Estonia because of the Baltic-German feudal order which lasted until the beginning of the 20th century. One can still presume that in the second half of the 19th century when the collecting of Estonian folk tradition mainly started, the informants themselves already tried to interpret parallelism consecutively. In spite of the centuries-long resistance, time parallelism (like any other kind of parallelism) tends to be eliminated from the ballad as natively unsuitable for it. Surely, time parallelism is notably lacking in the later end-rhymed traditional ballad. At the same time it is sometimes excellently imitated in later artistic poetry (e.g., the collection by Villem Grünthal-Ridala The Blue Herd `Sinine kari', 1930, and others).
And, last but not least - it is interesting to note that parallelism does not obligatorily appear in the representation of the present only. There is a version of the ballad Husbandslayer, in which the end part, making a reference to the future, is also represented in the double form. At first, the female protagonist drowns and then she is burnt. So the model H - P1 .... Pn - F1 .... F2 or some other version of it is also possible. Obviously parallelism can appear in the representation of the past, too (see the beginning part of e.g. Grave's Girl in this respect). Still that solution is rarely met in texts.
I have considered this fact on the draft, too.

fig2
Leaving out the (balladic) legends about Jesus and Mary, Estonian folklore evidently does not use this model. The historical time as tense is lacking in the Estonian language, and therefore the Estonian mind can hardly perceive the possibility. Nevertheless, the ballad system of the historical time is abundantly represented in Estonian literature (obviously because of German and perhaps also some Russian influences). The most well-known ballads, especially dramatic and romantic ballads of Estonian poetry, represent that system in which the presentation of a particular single event gets its interpretation and meaning from the more general time horizon. That time horizon suggests a more fatal inevitability. The single event under discussion is only a tiny particle on the background of the higher historical game which frequently can be even a supernatural one.
Ballads like those were written by Jakob Tamm and Jaan Bergmann in early Estonian literature at the end of the 19th century. In the 20th century they are represented by Henrik Visnapuu (the collection Mother of Winds `Tuule-ema', 1942), Karl Eduard Sööt (Crescent's Blade `Kuusirbi õsu', 1937) and some other poets. The highpoint of Estonian literary ballads, the collection Eclipse of Happiness `Õnnevarjutus' by Marie Under, also belongs to this model. Concerning that collection and the years when it was created, I have published the following thoughts earlier:
In the 1920s the ballad production becomes more active. /---/ A general lyroepic situation comes into being /---/. Especially productive are the years 1926-1931 (1927-1930). As a certain parallel one can notice the dramatic conflicts in society during the Great Depression, from where much of collision flowed into literature. /---/
M. Under's Eclipse of Happiness (1929) is the masterpiece of the Estonian poetry. /---/ R. M. Rilke's concept of Weltinnenraum /---/ obtains in Under's poems the form of a circle, its symbol being a ring or a round lake. The integral whole is formed of two sides, the man and the woman, who found their fertility on passing through a death-containing psychic space. If one of them sins against love on their way, the satanic powers will spoil the whole - a material sin will follow the mental one. In the opposite case happiness can be gained, although with reservations. The man and the woman sin equally five times, this does not befall only in the Bible-based Mandrakes. The lyric poem Swamp Song symbolizes the mental morass as a clue. The basic scheme of a Gothic cathedral is being formed: side-naves and prop-arcs as its body, portal and main way, pulpit (position of author), altar and belfry:
			    	Mandrakes
   	Young Lady of Porkuni			Leather Merchant Pontus         
	Child Killer        			Hobgoblin
	Whirlwind        			Sea Cows
	Exchanged Child      			White Bird 
	The Birth of Naissaar Island		A Travelling Lake                              
				Swamp Song

Just as in the Gothic, Under's stories reflect in details each other and the integral whole. It seems that through M. Under's Eclipse of Happiness the Gothic ballad has reached its historical closure, the absolute comprehension of its essence (i.e. the dramatic conflict between the lyric and the epic as a cathedral surging up to altitudes). /---/ With this collection of ballads her talent achieves perfection (Merilai 1991:138-139).

The equivalent to Under's ballads in prose of the same period are the erotic short stories by the Finnish-Estonian writer Aino Kallas. She has named her short stories "ballads in prose" and all of them present the same model of time.

fig3
Understanding this type is easy. This model may express the ballad in the most typical way. It is the presentation of a story in the most exact sense of the word, the original expression of the run of time. Many different examples can be found in Estonian literature through times. In folklore, the most definite example of that model is the later end-rhymed country dance ballad or market song. However, the ballad, pursuing even more dynamical presentation of the (maybe sensational) story, can even break off from that model as can be seen from the following example.

fig4
a) Time system of the nucleus of the story
In Estonian poetry some ballads have an extremely concentrated content and form, for example, Killevere-Kullevere and some others in the collection of poems Bird Song `Linnulaul' by Henrik Adamson (1937). The most interesting example is evidently Virgin Forest Lake `Põlismetsa järv' by Karl Eduard Sööt in his collection Home `Kodu' (1921). The ballad is close to Lorelei by H. Heine. The length of the poem Virgin Forest Lake is uncommonly short for a lyroepical ballad, containing only 2 stanzas and 8 verses. In spite of it the story has all features of a lyric-epical and mystical ballad. I have characterized these examples once as dramatic ballad fragments. Although Adamson and Sööt are very folksy in those stories, similar ballad fragments have not been found in folklore. They are just the products of literary condensing.

b) Time system of the emotional background of the story
In my mind, the system of in maxime medias res can be seen also in the poems which represent the story quite fragmentarily and in a concentrated way, but in which the poetic attention is not so much directed to the dramatic nucleus of the story as to the presentation of the emotional background in the wholly mystical manner of the ballad. Such are, for example, the ballads of pantheist and symbolist visions by Ernst Enno in the collection Grey Songs `Hallid laulud' (1910), and especially the poems by Bernard Kangro in the collection Drying-Kiln `Reheahi' (1939). Here belong also the ballad fragment type stories by Adamson and Sööt.
The in maxime medias res system is a grey area of transformation between the ballad and common lyric. In this case the ballad is as if it were dispersed into more primitive components: either into the presentation of the story in a minimalistic form or into the expression of the emotion caused by the story. In longer ballads both occasions would exist together as a compound. I have characterized the Kangro version elsewhere as follows:
A remarkable ballad wave appears in 1937 and in the following years. /---/
B. Kangro's poetry is close to H. Adamson's, but at the same time intellectually opposite to it. The traditional ballad he experimented with in the collection Old Houses remained alien to him. He makes the genre more lyric, stresses the unity of opposites instead of their antagonism, and as a result the plot becomes more fragmentary and dramatism is replaced by a dreamlike condition (Drying-Kiln, 1939). As a matter of fact it is ballad-like poetry with obvious distinctive marks of the ballad: mytho-poetry, an urge for more ancient mind-strata than the medieval man's. /---/ Kangro's theme, as in all the best Estonian ballads, is that of the whirlwind of the soul in the grip of external forces (Merilai 1991:140).

This model prefers psychical time to real time, whereas in the case of Under both times were equal.

fig5
The system seems to work when the purpose is not so much to express the content of the ballad as such and through it the time as such, but just to write a ballad as a form which has become creditable. The secondary time system imitates some previous system, but it itself is not concentrated on the expressing of time. The purposes of a system like this are literary, formative. The writer strives for the balladic artistry. The attention is centered on the manner of presentation, the literary allusions and the "genre memory" problems, but not necessarily on the metaphysical basis. The expressing of time may be "forgotten", but as a matter of fact it can never be amiss, as the world and the stories in it are temporal. In that case it is as if the ballads were not genuine, but rather the poems or ballads about ballads. They are more genre reflexive than time reflexive poems.
In our literature the elements of the secondary system can be met rather early, but the system activated at the end of the 1930s. In my mind it can be noted in the best ballads by Betti Alver and Kersti Merilaas already. The "timeless" system is typical of the 1960s, when the ballads were usually written in vers libre form. The vers libre form was the purpose then, the rest being of less importance. When a writer chose the ballad form, the choice was not made to express metaphysical time. The time was only an unavoidable background feature. The valuable poems written in this system are by Ain Kaalep, Arno Vihalemm, Jaan Kross, Paul-Eerik Rummo, Jaan Kaplinski, Mats Traat, Aleksander Suuman, Lehte Hainsalu and recently by the young poetess Kauksi Ülle, who writes in south-eastern Estonian dialect.
This system promoted the ballad to the level of cultural refinement. Not every writer was able to meet the requirements presented by the level. So the ballad could easily become its own parody or a pastiche expressing only non-cultural non-history instead of time. The ballad could turn to a tasteless imitation or a kitsch. By the way, during the period of Soviet occupation, too, people were suggested the mentality of historical completion and therefore timelessness. The physical side of time was rendered too important, the metaphysical side was ignored. Unable to find a way out, the kitsch writers accepted the misleading mentality. However, this might have been a step forward from the cabaret style of Berthold Brecht. In this system the ballad looses its metaphysical independence and turns into the tool of literary games. In this case the poetic role of the ballad becomes replaceable salva veritate by sonnet, poem, haiku, gazel, unrhymed poetry or any other form, even by an article on the current literary life in a newspaper.

5. The system of independent time units
 fig6
Not all Estonian poetry (as well as folklore) is synthetic, so it is not totally balladic either. Estonian poetry can contain also time-expressing systems where different time segments can, but need not form a unified linear string. Thus they are analytically independent. An example of the analytical poetry is the conversative aphoristic unrhymed free verse poetry by Artur Alliksaar in the 1960s. While in synthetic poetry the orientation act is committed on the level of the discourse as a whole, in analytic poetry the orientation is committed separately on every utterance level, whereby the time deixis of the utterances, following each other, can be discordant instead of being in accordance.
In the situation where each utterance or group of utterances can code a different context, the expressing of linear time on the level of discourse may turn problematic. Different utterances can express different time which may be incoherent. The time structure of such utterances may be incongruous. The time deixis of the text as a whole is more like a plural conglomerate of separated time icons, defined by different utterances. (Ballad as such presents only a single icon of time.) It is possible to play various games and create paradoxical situations, using the time deixis and implicatures with it. An icon can represent the common time sequence in time (like in ballads), but also the movement from the future to the past. It can be internally complete but also defective, and so forth. It is frequently characteristic of Alliksaar that the time is paradoxically concentrated into cairoses, where the past, the present and the future fuse together, loose their identity and make up an ecstatic time of "overtime". As a rule, then Alliksaar replaces the solipsist first person singular me-origo by the extremely inspirational we-origo. The we-utterances with their multi-concentrated centres of orientation (at most ellipses) are characterized by notably higher level of mental co-operation. In the poetry by Alliksaar the co-operation is performed actively on behalf of a fictional group of Argonauts or a special kind of Pickwick-Club of intellectual aristocracy.
In the poetry by Alliksaar, the analytic time moments can make tiles or parquets. Symbolically it could be described as follows: in which i, j = 1 .... n, graphically as follows: 
T1    T2       ....   
 Ti        ....       Tn
The moments can also make cascades in which the intersection of the time sets is not empty:
in which i = 1 .... n - 1.

Time icons can localize (quasi)parallelistically against the background of one-to-one correspondence, but the accessibility relation of one time from another is not necessary. The parallelistic equality condition of time moments is not unavoidable either. This case can be presented by the model of a pack of cards:
fig9
To sum up, in the analytic poems by Alliksaar time as the metaphysical basis of the text does not move linearly. Time is plural, multi-directed, restricted in different possible worlds. Alliksaar has also written traditional synthetic poetry, e.g. Lemon Ballad `Sidruniballaad' for children which belongs to the V system. The ballad does not start in the VI system, but it can be realized in some subdivision. Some balladic fragments or their parodies in the boundaries of some single time icon can appear in those texts. (Alliksaar is ironical about everything except his mental club, not specially ballads, but what's wrong with irony?) Although it is not the writer's special wish, those poems could not be ballads at the same time, even if one wanted them to be, or we should re-define the essence of ballad as a synthetic form.
Thus the ballad is not metaphysically so universal and uniform as I supposed earlier, although its is remarkably prevalent and highly influential in common poetry. Even if the ballad as a form should become anachronistic, serious efforts have been made during centuries, so that the final disappearance of the balladic time is impossible.
Dixi et animam levavi.

References

Cheesman, Tom 1995. E-mail to A. Merilai in July 12.
Currie, Gregory 1990. The Nature of Fiction. Cambridge: Cambridge Univer sity Press.
Forbes, Graeme 1985. The Metaphysics of Modality. Oxford: Clarendon Press.
Green, Keith 1992. Deixis and the Poetic Persona. - Language and Literature. Vol. I, No. 2, pp. 121-134.
Grice, H. Paul. 1989. Studies in the Way of Words. Cambridge, Massachu setts: Harvard University Press.
Merilai, Arne 1991. Estonian Ballad 1900-1940. Summary. - A. Merilai. Eesti bal laad 1900-1940. Tartu, Tartu Ülikool.
Kurrik, Juhan 1985. Ilomaile. Anthology of Estonian folk songs with translations and commentary. Maarjamaa, Toronto. Studia Estonica Fen nica Baltica. Vol. III.
Searle, John R., Vanderveken, Daniel 1985. Foundations of Illocutionary Logic. Cambridge: Cambridge University Press.