From the WAB archives: A selection from the Bergen Wittgenstein archives' Working Papers and audio-visual materials

• Home
• About
• Login
• Search
• Current
• Archives

Table of contents

• References

The aims of the Wittgenstein Archives are neatly circumscribed. The work is, however, taking place within a vast area of research, constituted by several different fields partly overlapping each other. A few of these fields, like textual criticism, have a long history, others, like the development of software for different purposes, are a result of efforts made during the last few decades to systematize theoretical and practical problems involved in the use of information technology. If we lump the relevant research fields together by the designation "textual studies" – or even "textual sciences", a term apparently growing into fashion – we may discern certain fundamental problems common to the most diversified disciplinary fields within this large area. In commenting upon Claus Huitfeldt's thought-provoking paper, I should like to point out some problems of particular relevance for future discussions.

The implications of "machine-readability" within textual studies range from very practical problems in the construction of databases to complicated speculations on the effects of using non-linear texts, like hypertext (programmes giving the possibility to browse in large masses of material and to construct different textual entities). The international argumentative situation constitutes to some observers a promising and to others an exasperating condition of "betwixt and between". I borrow this characterization from the anthropologist Victor Turner; it is used by him to describe the function of rituals in sacralizing contexts of meaningful action by setting them apart from profane everyday life. Rituals thus organize certain interpretations of changes in social life.

Some rituals of the scholarly world are certainly undergoing modifications, resulting from the strain of adjusting to the demands for as well as the needs of institutions like the Wittgenstein Archives. New coalitions and new factions pop up – not in the manner of fungus, rather more often than not as a result of cumbersome cultivational work. The staff involved in the real research and footmanship of this kind of institutional upstarts within the University society has no reason whatsoever to ask for a more guarded conservatism, neither in bureaucratic policy nor in common scholarly modes of thought, than what has hitherto been displayed. It seems deplorably safe to predict that for some time yet to come, computerized text analysis or information technology within the humanities shall not represent major challenges to the formalized conduct of (Norwegian) University budgetary and organizational measures. Like ritualized behaviour generally, administrative measures tend to incorporate new definitions of themselves and their relations to the surrounding world in such a way as to further mythical explanations of practical problems, without bothering to worry about flagrant discrepancies between what people are supposed to accomplish and what means they are given to accomplish.

No, the real challenge represented by the automatization of textual criticism should be sought elsewhere. It should at least partly be sought in the possible de-automatization of some scholarly conventions for dealing with textual material.

I venture to suggest that this "elsewhere" ought to be properly mapped. Furthermore, that it should be a proper task for institutions like the Wittgenstein Archives to contribute to such a mapping. By the term "mapping" I refer to procedures that are mostly underestimated by laymen (in our context that is, scholars observing any field of research they arent originally initiated to), notably the efforts of getting to know the proper questions to ask in order to get to know the landscape you are moving in. There is need for a topographical registration of the cultural geography of these intellectual fields, where viewpoints and itineraries of the people moving there differ to a considerable degree, according to their positions as well as to the way they use their particular field maps. Such "topographical description" is a phase of research often overlooked in the humanities, which accounts for much curious eclectisism as well as for some rapid changes of scholarly fashions in shirts and arm-chairs.

Generally, the consequences of suchlike scholarly ignorance are that new theories or methodologies in the humanites tend to be met either by a dull so-called "sound scepticism" without much real scientific discussion, or by enthusiastic great expectations bording on fantasy. (After some time, the state of affairs within this argumentative discourse tends to create a certain melancholic frustration in apprentices as well as adepts suggesting new solutions to a problem.) On the question of machine-readable versions of different kinds of textual material, the attitudes towards scientific problems involved tend to fall into either one of two major categories on each end of a scale. One end of the scale belongs to the waiters of traditional disciplines ("Sorry sir, this is not my table"), whereas the other end is represented by the entrepreneurs of new enterprises ("This is the future, imagine the profits, man!"). Somewhere in between, or perhaps altogether outside such a scale, we find some knowledgeable people telling us about the actual practical limits and the theoretical puzzles involved. Claus Huitfeldt's paper certainly places him and the Wittgenstein Archives in such a tertium datur-category.

Huitfeldt himself does contribute to a topographical description of problems within the research field, by describing practical problems and by pointing at the necessity of reworking traditional concepts of textual analysis as well as the philosophical problems evoked in this process. This is altogether another attitude than that of the above mentioned waiters and entrepreneurs of the humanities; neither content to serve nor to rule the field, the Wittgenstein Archives strive to redefine the framework for the coexistence of different fields of textual studies.

Trying, now, to restrict myself to the subject at hand, notably the transcription and editing of Wittgenstein's texts, I should like to point at certain areas of questioning. I believe that four areas are in particular need of being discussed:

1) the relation of philosophy to philology,
2) the problem of intentionality in texts,
3) the relations between orality and literacy, and
4) the possibility of a critical hermeneutics.

Consciously, I have chosen to divide my comments into four subsections, in order to avoid the mythical triade. The division in three parts of a story, a lecture or a paper is so often used for pedagogical and argumentative purposes that it tends to be taken for granted as an organizational principle of academic discourse. But beware; the triadic division also tends to evoke the fairy-tale impression of a well-rounded and already finished process of thought, as produced by the narrative scheme of beginning, middle and end.

My point of departure is an aphorism by Wittgenstein, found in the notes published as Philosophical Remarks. I shall elaborate my understanding of the point made in this aphorism before I proceed to my four areas of questioning. Like so many others I have found the reading of Philosophical Remarks rewarding, the volume bears testimony to the gradual reformulation of argumentative strategies and perspectives in the transitional phase of Wittgenstein's thinking, in between the Tractatus Logico-Philosophicus and the Philosophical Investigations. The second period of § 66 reads:

Philosophers who believe you can, in a manner of speaking, extend experience by thinking, ought to remember you can transmit speech over the telephone, but not measles.

By the same token: You can replace the signs of more or less coherent typescript or handwriting by sequences of codes that will be seen as offering information when represented as texts on the computer screen. But you cannot transmit the experience of having accomplished such a feat! You won't be able to codify the different kinds of feelings, thoughts, tacit knowledge or even all the overtly conscious knowledge that goes into the work. In order to even try to do such a thing, you will have to construct another text. And in order to explain what went on in your work on the explanatory text, the "text of second order", you'll probably find it necessary to write still another text. And so on, in an infinite regress.

A comparable problem is constituted by so-called "metapoetry", the comments directed towards the art of writing or reading poetry or fictitious texts that might be included in the text or even made into the major theme or motive. A writer is not obliged to put into his text explanations of how and why he came to write the text in the way he wrote it, and the general public in our time doesn't expect a poet or a novelist to do so, either. Still, he might do exactly that, and by subtle hints or by explicit postulates demask his interests, his dispositions or his position in the world – in the way modernists and before them baroque poets incorporated the Sitz im Leben of the writer.

Diversified metacomments, for instance in the form of allusions or allegorizations on the writing process or the function of the text, most often mark the poem or the fiction as non-realist. Meta-poetry tends to destroy any illusion of the text being a direct report on occurences in the world, and implicitly problematizes the function of language.

Such metacomments should not be regarded as reports on the poets' experiences in writing poetry or creating fictitious universes, even though the very term "metapoetry" invites us to think that the author is stepping out of his text in order to comment upon it. This view of the text is in itself metaphorical, it tends to stear the reader away from the question of how such constructions of a writer's perspective become features of the poem itself. These features are part of the very "texture" whose genesis or function or tendency (etc.) could be explained and which, therefore, is the result of reflection upon the experience of reading the poem.

This kind of reflection cannot possibly double or imitate the experience of the poet in writing the poem, even if the poet's intention might have been to invoke particular kinds of reactions. This is to say, contrary to popular belief, that a poem does not transmit the experiences of the poet to the reader, even when the poem itself gives clues to the principles of its own creation, and it does not fix the experience of the poet in writing. Popular belief in how the impact of poetry comes about is a variant of the positivist view; one imagines that the poet calculates his "signals" as if they were arrows hitting the reader's sensibilities, to make the proper reflexes and sentiments surge forth. But this kind of vision tends to overlook how reading is an activity steeped in and dependent upon other kinds of experience. The study of poetry could be said to represent a special case of more general investigations of how awareness of experience comes about.

Traditional hermeneutics used to consider texts as an expansion of the reader's field of experience in the "encounter" with the text, whether the concern is to reconstruct the original "meaning-content" or to explain the "reader's response". In some respects, the deconstructionist trends in textual analysis relie upon a similar notion of how texts affect the experience of the reader: preconceived notions, propositions presupposed or explicated in texts, figurations of imaginational figments or illusions are broken down by contradictions created by the reading of the text. Strangely enough, the very basis for discovering homologies between textual material and experience tend to be overlooked: both are constructions.

The tendency to treat "experience" as if it were something purely given – individually or collectively – goes with the view that "reality" is something originally pure and the experience of it is "tainted" by ideology, theory or bias. Descriptional procedures derived from phenomenology, hermeneutics and deconstructivism included, tend to perpetuate the dualism between "real" life and (illusionary) "constructions". The ensuing discussions of "realism" as opposed to "relativism" in their turn tend to obfuscate the subject matter by diverting into scholastic quarrels on how to define and thereby confine the opponents' positions as well as one's own.

Instead of clinging to the age-old antagonism between "real" and "constructed" one might hold that the experience of reality is constituted by constructions of competing descriptions that are given explanatory force. This is to say, nothing more and nothing less, that our experience of reality is dependent upon our perception of conflicts between different ways of describing the causes and effects of changes. If we are aware of such conflicts, we may see options and strive to make a choice, if not, we accept the received views and stand or fall by them. Becoming aware of conflicts or possible conflicts in the description of life implies on the one hand to start to question the construction of descriptions, and on the other hand to face the challenge of constructing our own description(s).

Accordingly, the said antagonism between "real" and "constructed" dissolves. It may be replaced by questions of how the processes of construction of experience and the construction of textual understanding may be compared to each other and how they interweave.

Huitfeldt's paper wisely calls for reforms in textual studies, and poses some challenges to certain presuppositions amongst literary critics on how one should treat the problems of textual criticism. The paper does so by making classical problems in textual criticism as well as new methods for tagging texts relevant to the development of philosophical problems. It might hold some interest to elaborate somewhat on one of the traditional concerns of textual criticism, the problem of establishing a textual stemma, a necessary prerequisite for any kind of textual criticism that compares older and newer versions of texts.

The philological method stemmatology is the reasoned construction of a tree-like model showing how new copies or versions of a text are branches on a common stem. A stemma makes for a model, or a map, of "family resemblances" for texts with common origins ("stemming from somewhere"); these common origins show by correspondancies and dependencies between different text versions, so that it may be possible to reconstruct a history of different schools of copists, or, in certain cases, different traditions of dealing with the subject matter (for instance, different narratological patterns). Such is also the general problem outlined by Huitfeldt for the special case of describing the development of Wittgenstein's writings. The description is intended to organize the text material in certain manners, so that programs for machine-reading makes the search for an origin, a line of development and the comparison of different versions easier and more efficient. This is, by the way, congruent to Wittgenstein's own view that description has a logical priority over analysis.

Models that systematize stemmas to make tagging coherent are, like all models, replacements for that which is modelled. Such models fill similar functions as theories do in relation to the objects studied, but the assumptions made in the descriptions aren't always explicated. Models represent systems of relations between elements, thereby subsuming different elements within categories. This is also the job metaphors do in the use of language. They tend either to uphold old categories or to create new ones, and thereby alter classificatory schemes; the metaphor gives a focus, a way of regarding objects, themes and problems. In a similar way, a model prescribes how objects or problems under study should be regarded.

Stemmatology and other kinds of philological reconstructivist procedures have old practical uses in different disciplines. The construction of a stemma has been employed in order to date manuscripts and in certain cases printed versions of a text (in particular, incunabula), or to reconstruct an original version of a text, eventually suggesting a starting point for diverging manuscript traditions.

One of the more famous examples of how stemmatology has been put to use, is the conjecture (not reconstruction) of a Gospel of "Q", a lost text supposed to be the original main source (Q = Quelle) for the three synoptic gospels of the New Testament. Such conclusions are reached by analyses of structural features of the text, "Gattungen" (genre features), markers of "Sitz im Leben" and topics dealt with in the synoptics, in this case paleographic evidence from different manuscript versions has played a lesser role than in most stemmatological reconstructions. In other cases, external evidence from archaeology or even references in other texts have proved to be important. Related problems are constituted by the text material found in Qumran in the period 1947-1956, the texts popularly known as the Dead Sea Scrolls. To the detriment of many a scholar of ancient Middle East culture, history of religions, biblical studies etc., the major part of the Qumran texts never have been published or been made available for normal research. Thus suspicions arose, and in recent years blatant accusations, that those in charge of the Qumran material willfully have delayed the publication as long as possible and that their published interpretations of certain texts or text fragments are deliberately misleading, furthermore, that the reason for these machinations is to prevent a possible fall into disrepute of "official", that is, Church-sanctioned, theological explanations of how the Qumran material affects interpretations of New Testament texts. When all relevant texts finally are made available for philological scrutiny, it will be possible to construct sound procedures to replace the hitherto more or less impressionist methods for dating Qumran texts, as well as the rather haphazard conjectures constructed to explain their context. New programs for machine-based analysis are bound to become the focus of attention in a not too distant future.

The Qumran quarrels is only one among a host of examples of how philological research in our time integrates a critical examination of handed-down beliefs, or "doxological systems". The Wittgenstein Archives could be regarded as another such example of how older notions are being reevaluated. The Wittgenstein Archives could, possibly, also become a test case in establishing sound principles of evaluating doxologically bound traditions, that is, accepted canons of systematized interpretation, within philosophy as well as philology. The Wittgenstein Archives prepare a well-defined material (produced by one individual only) for machine reading. But by doing this work, they also prepare the grounds for wide-ranging future discussions on the interpretation of Wittgenstein's work, as well as on the relations between philosophical and philological investigations.

The leap from Qumran and the New Testament to Bergen and the Wittgenstein Archives is not that farfetched. Again I shall allow myself to get to the point by the somewhat roundabout way of another historical example. "Comme chacun sait", as the French say whenever they feel that an audience ought to be reminded of what they should know, the starting point of early modern textual criticism was the exposal of a forgery. The problem of forged manuscripts, as well as the subcategory of plagiats, has represented an intellectual and moral challenge to philological disciplines ever since the celebrated Italian humanist Lorenzo Valla attacked the supposed "Donation of Constantine" in the first half of the fifteenth century. It was the intimate knowledge of Latin usage of different epochs that made possible the most important features in Valla's deconstruction of this letter, which stated that the Emperor Constantine donated the Lateran Hill to the Church. The authenticity of this document had been questioned before, but Valla analysed the linguistic forms as well as certain anachronistic descriptions, such as terms referring to ecclesiastical positions that didn't exist in Constantine's time. He concluded that the "Donation" had been written in the early part of the twelfth century. Thus Lorenzo Valla established the basic principles of textual criticism. Ever after, the examination of textual authenticity has been an integral part of philological research.

Somebody might think that the mention of Valla's demasking of a mediaeval forgery could be a kind of unsubtle hint at a registered dissatisfaction with the actual condition of some of the published Wittgenstein works. Somebody might be right. But this is not to say that any of the published Wittgenstein works are tampered with or falsely reconstructed – as were the case of Niet[zsche]'s Nachlaß, for instance. I just want to point at the obvious possibility that in some cases editors have been guided by particular interpretational perspectives in their choice of how to render different versions of Wittgenstein's original manuscripts. Whereas this is a quite honourable and has generally been considered an acceptable practice, the problem remains that the published texts by their form and selectiveness constitute the availability for the reader to a certain set of possible questions. The organization of the text material favours certain kinds of interpretations.

One rather obvious question to ask, is: during the process of making the authentic versions of Wittgenstein's texts – whether they are fragments or interpolated manuscripts, available to a general academic public in the form of machine-readable versions – what might the Wittgenstein Archives do to develop discussions about the interpretations presupposed in already published versions? And the most obvious answer to this question is that the Wittgenstein Archives cannot but leave these publications in their present state, but may nevertheless contribute to an evaluation of them. The less obvious question to ask, perhaps, is: what might the Wittgenstein Archives do to focus the question of how biases in interpretation are affected by the new kind of availability of Wittgenstein manuscripts? This question may be rephrased to give it a more general scope: what might be done to further the awareness of relations, or rather affiliations, between philological criticism of texts and philosophical understanding of texts on the basis of this particular collection of manuscripts?

I quite agree with Huitfeldt when he sticks to the pragmatic dimension, without launching a quest for ontological definitions. Such questions as "what is a text?" tend to create too many impractical definitional boundaries for the concept of text. A pragmatic approach to the understanding of texts is to stress the question of how rather than the question of what a text is. The only feature common to every text is that it is a part or a specimen of discourse fixed in writing by conventional signs. I do insist that spoken discourse, video recordings, films and even untranscribed tape recordings should not in their primary manifestations be regarded as texts, even though such material may be analysed by the same or similar methods as those employed in textual analysis, and even though they might be transformed into texts.

The reason for using a minimal definition is that it concerns the means by which discourse is fixed – by writing – and therefore poses particular problems in the understanding of how writing and reading functions. It might be that the proposed minimal definition of text is somewhat fuzzy. I tend to regard this kind of fuzziness as an advantage. It gives the possibility to search for principal problems in understanding the practices of writing and reading, whereas starting off with the grand question of how to distinguish essential features of "the text itself" tends to make us jump to conclusions. The ontological question doesn't allow for the study of how a text is formed and how it transforms the subject matter.

Philosophers are keen readers, trained to discern the argumentative possibilities given in a text, often with a very high degree of sophistication. Unfortunately, quite a few philosophers also often seem to be oblivious to what goes into the constitution of the texts they are reading. This forgetfulness seemingly results in a desire to leap directly at the Sachlage, the theme or subject dealt with in the text. This desire to make the grand leap is manifested in the theories of hermeneuticians like Gadamer and Habermas. Gadamer, in his Wahrheit und Methode, explicitly relegates the questions on form to the realm of technicalities and methodology, and considers the search for die Sache to be the real philosophical quest. Habermas, in his Theorie des Kommunikativen Handelns, systematically overlooks any kind of linguistic and philological evidence to counter his grand theory of how civilization developed, and brushes away rhetorical theories of language use as being irrelevant.

By the Gadamerian and Habermasian moves, philology (in a broad sense, textual criticism as well as literary criticism) is relegated to the position of philosophy's hand maiden, preparing the field for the real action. One could imagine worse positions, of course, and since the middle of the eighteenth century, when philologists became selfconscious and taken up with the task of defining the scientificity of their endeavours, they have usually been content to forsake what they regarded as metaphysical speculation – with some major exceptions, like the speculative work of Herder and the later historicists. After the time when Schleiermacher had laboured towards the unification of philosophy and philology by the joining of hermeneutics and rhetoric, a new divorce process was conducted by Schleiermachers pupil Boeckh in the 1880'ies. Ever after a mutual suspicion has strained the relations in these disciplines' cohabitation within the faculties of arts. The genesis of these uneasy relations may be followed backwards in history to the humanist quarrels of the Renaissance, when Lorenzo Valla, in his usual sarcastic mode, characterized the practice of philosophers, "who", as he put it, "restrict to themselves the name of friends of wisdom". Well, on the other hand it might be justified to say that since the late eighteenth century philologists often have been too happy to remain "friends of words".

This division of labour between disciplines, and the division of world-views it entails, becomes untenable when one considers the implications of preparing a critical machine-readable edition of the writings of such a philosopher as Ludwig Wittgenstein, in particular when the code system (as described by Huitfeldt) is designed to serve multivariate purposes.

Doesn't this very classical philological work of reconstructing and reflecting upon the many different versions of Wittgenstein's formulations, examples, aphorisms and elaborations of thought patterns imply a challenge to certain philosophical presuppositions about how ideas, thoughts and arguments are developed by their mediation in writing?

I now introduce my second area of questioning, that of intentionality. Doesn't the transcriptional work in itself beg the question of how to get at, describe or understand the author's intention? Transcription is not so menial a task as some would have it, but implies serious considerations on what is more and what is less important, and careful employment of signs, codes and explanatory devices.Therefore, transcriptional work also implies a question of how to interpret the author's intention.

Wittgenstein claimed, in Philosophical Remarks, § 20: "If you exclude the element of intention from language, its whole function then collapses". §§ 23-24 read as follows:

§ 23: If, when learning a language, speech, as it were, is connected up to action, can these connections possibly break down? If so, what means have I for comparing the original arrangement with the subsequent action?
§ 24: The intention is already expressed in the way I now compare the picture with reality.

"Picture" in this context must be related to the kind of visual pattern that Wittgenstein in another context calls "paradigm", when discussing the appearance of contradictions in mathemathical calculations. I quote the very last periods of F. Waismann's shorthand transcript of Wittgenstein's talks and conversations 1929-1931 (placed by the editor as Appendix 2 in Philosophical Remarks, Oxford 1975 edition, pp. 317-346):

What are we supposed to be afraid of? A contradiction? But a contradiction is given me with the method for discovering it. As long as the contradiction hasn't arisen, it's no concern of mine. So I can quite happily go on calculating. Would the calculations mathematicians have made through the centuries suddenly come to an end because a contradiction had been found in mathematics? If a contradiction does arise, we will simply deal with it. But we don't need to worry our heads about it now.

What people are really after is something quite different. A certain paradigm hovers before their mind's eye, and they want to bring the calculus into line with this paradigm.

To start the reasoning backwards: This last statement seems to imply something very much alike the point that Thomas Kuhn later was to develop into a criticism of Karl Popper (especially in a few of the essays in The Essential Tension): It is simply not true that scientists normally adjust their way of reasoning to the experiences they gain in discovering falsifications. On the contrary, scientists tend to force their findings into the paradigm they have adjusted to, that is, to the way (Greek: methodos) in which they are used to deal with a problem. Therefore we have problems in analysing real change when writing the history of science, as opposed to the linear description of how one solution to a problem neatly fills the place of an older one.

Such paradigms, or exemplary ways of dealing with a problem, constitute objects of "the second order" for research, of how to evaluate the outlook of the analyst. This brings me back to the question of how already made interpretations (or ready-made interpretations) and presuppositions (or biases) determine the edition of textual material of this kind. The point I am trying to make is not that editions are "tainted" by the editors' biases, but that the critical examination of scholarly presuppositions is a necessary prerequisite for assessing the value of new suggestions. And that such an examination has some rather intriguing implications for the analysis of intentionality.

It is a question of how "the picture of Wittgenstein" in "the mind's eye" of the editor affects both the possible intentions that an editor may recognize in himself, and the recognition of possible intentions in the writings of Wittgenstein. On the one hand, this is to say that the intentions of an editor aren't entirely his own, but that he in the very process of acquiring intentions and putting them into action by necessity must depend upon his training, his intellectual field of work, his position in this field, his colleagues (perhaps) or his public – and various other elements that might go into the construction of intentions. On the other hand, this does not amount to saying that the intentions of the editor (or the in casu text encoder) does not belong to himself at all, because he takes on responsibility for the intentions he puts forth in acting – like the action of editing or encoding. So, intentions belong at one and the same time to the acceptable and recognized moves in practice (in the same way as concepts, according to Kjell S. Johannesen's paper, "can be regarded as a function of the established use of its expression") and to a personal sphere of responsibility.

This view of intentions as embedded in action – in this case the acts of writing, of editing writings and of reading texts – makes for another and somewhat more dynamic view on the structure of a text than what is often presupposed. The structure of a text has by necessity an historical dimension as well as a social dimension, the merging of these dimensions is usually shown by distinctive stylistic features, in particular the genre. The individual work within every genre (for instance a poem, a novel, a letter, a thesis, a treatise) is endebted to other texts within the same or even in other genres (a characteristic feature of an essay is that it quite freely may use elements from different genre traditions). We may say that such relations of dependance is the result of a process of adaption on the part of the writer, we may regard it as an application on the part of the reader, or we may label the phenomenon "intertextuality". The point remains that a text invariably show traces of the impact other texts have made – in phrasing, in the uses of signs, and – if it's not too small a fragment – in genre features. Some genres have a history of being anti-genres, for instance the novel from Cervantes to Diderot or modernist poetry, their "raison d'être" is a kind of protest against earlier forms of the genre and by this mode of existence they are transforming the genre. In a similar way, Wittgenstein's way of writing is from the outset negatively dependent upon the earlier way of writing philosophy, and he develops his style of writing into a way of philozophising against other styles and textual patterns in the process of establishing his own views.

The interpretation of the intentionality of a text is to interpret the possible frames given for understanding by the stylistic features of the text. We might say that textually structured intentions live a "double life"; that of the literate culture they are inserted into and are dependent upon, and that of the individual contribution represented by the text. Style as the result of conscious creation is opposed to accident, style is the individual way of the writer to impose his will on his medium.

The double life of intentions fascinated the ancient sophistic rhetoricians, who put their knowledge into more or less effective use by seeking to influence their contemporaries to choose between possible patterns of intentions. Unfortunately, Platonic philosophical tradition always tended to dismiss such choices on the grounds that to place people in such situations was an immoral way of trying to exploit people's faith as well as their bad motives. This moral pointer distorts the problem of understanding how we acquire intentions.

In everyday life we do feel free to say that this or that person doesn't understand what he is saying, or even to claim that X and Y doesn't know what they are doing. Such talk is, under certain circumstances, held to be quite literally true, as for instance by some of the participants in the discussion on whether or not Knut Hamsun, Nobel laureate in literature, really understood what he said when he condemned the dying KZ-prisoner and Nobel Peace Prize winner Carl von Ossietzky as a traitor, or what he later wrote in his necrology on Hitler.

Far less dramatic examples, and more intellectually intriguing ones, are given by different kinds of authors returning to their own text; revising it, dismissing it entirely to later extract and use parts from it, reappraising the text and interpolating sections to it or even commenting upon it. The point made by the old and worn joke that makes someone saying "I won't know what I have written until I've read it, will I?" comes to mind, a more pertinent comment could perhaps be "I don't know exactly what I meant until I have rewritten it". When asked about what he meant to say in this or that work of his, the novelist and poet Tarjei Vesaas used to answer: "I meant what I wrote". But he also is recorded to have commented upon the papers of a conference on his writings, that he had never before thought of many of the meanings now attributed to his works. Such a comment is not necessarily ironic, I should rather think it refers to a common experience for many a writer: the text not only takes on new meanings or loses some of its old meanings as the world changes around the text, but it also unfolds new kinds of possible meanings – or even new kinds of meaninglessness. What is at one period of time regarded as redundancy of meaning in a text, might later on be considered as constructive elements or destructive clashes of contradictions.

And thus we arrive at the problem of "the first order" when asking how intention is established in writing, notably the question of how to understand the intention of the original author who wrote these texts that are being edited, encoded, regrouped and generally put under the scrutiny of textual criticism. Please allow me repeat the quotation from § 24 of Philosophical Remarks: "The intention is already expressed in the way I now compare the picture with reality". Note that it is the question of how the comparison is made, the methodical features of the comparison, that is underlined.

What is, then, to be compared in the process of transcribing these particular texts? Different versions of the text, one would presume. Wittgenstein's collected Nachlaß constitute a rather bewildering mass of reformulations, cut-up fragments, fragments pasted up in new orders, interpolations, emendations, slashes and almost any kind of variations on the original versions that one might dream of – or have nightmares about, I should think, if one is charged with the task of organizing this strangely amorphous material.

Supposedly, a certain temptation to violate the form – or rather, formlessness – of this given material might present itself to anyone approaching it. It is constituted by the impatient desire to reformulate any how-question into a what-question, like: "Oh, bother, what is Wittgenstein's aim here? What is he talking about? What is his subject?" Of course, a concordance (as, for instance, arranged by a programmed search through the files) may relieve the reader of most of an elsewise painstaking job of contextualizing expressions, concepts and subjects. Working along the lines of identifying "key words" is to choose a method of reading that mediaeval exegetes used to call "lectio brevio potior", representing the view that the simplest interpretation is the best, we may somewhat inexactly translate the scholastic slogan into: "the best way is a shortcut".

The resulting new conceptualizing of Wittgenstein's manuscripts might lead to perspicuous analyses of his concepts. The essentialist temptation in such cases might, however, be to recreate Wittgenstein systematically as a kind of concept-maker, a filigree word-smith hammering out his conceptual universe, one tiny bit after another. One would then, in a way, have access to "the way I now compare the picture with reality", that is, if one really presupposes that Wittgenstein's concepts and ways of expressing himself are to be considered pictures of reality. And one has then made the feat of reconstructing Wittgenstein into his very opposite. – I mildly suggest that in the adoption of essentialist procedures resides at least a certain danger of misrepresenting the material.

A more exciting and comprehensive approach to this bulk of manuscripts is, I do believe, to adopt the competing scholastic principle of "lectio difficilior potior"; the most difficult reading is the best. We should accept the very formlessness – apparent or actual – as an invitation to examine how Wittgenstein conducted his own investigations into his own philosophical language. By comparing different versions, could these versions be said to have different clues to understanding, could their patterns of construction be discerned to be distinctively different in any way? Could rhetorical points be differentiated by their stylistic or aesthetic features, and do such features have determinate functions in Wittgenstein's recorded second thoughts and reformulations? In particular, what could be done to give access to a comprehensive study of Wittgenstein's examples? The way this philosopher stresses and elaborates on different types of examples is a constitutive part of his philosophizing, it is probably the one most important methodical feature of his argumentative strategy.

Marcus B. Hester discussed, in The Meaning of Poetic Metaphor (The Hague/Paris 1967), how figurative language constitute ways of seeing. It is the stylization of the metaphor that defines what possible intentions could be discerned. Hester underlines the Wittgensteinian point that intention must be defined by the way in which the poet expresses himself. His considerations are based on Wittgenstein's understanding of language use, accordingly, Hester's views could hold some interest when we discuss the representation of Wittgenstein's own uses of language. The use of examples and counter-examples seems to constitute Wittgenstein's own critical method. The Nachlaß shows how Wittgenstein incessantly returns to his examples, to modify them and comment upon them from new angles. Contrary to what is often taught, Wittgenstein does not usually "use" examples as practical illustrations for some more or less abstract principle. Instead, he explores the possibilities for thought in the use of examples, and is always on the lookout for such principles as are embedded in an example. In some respects, Wittgenstein's treatment of examples resembles the kind of new uses an artist might find for an "objet trouvé": The impression is sometimes given that he has found the example, and then asks what it might be an example of.

Wittgenstein constructs his arguments enthymemically. That is, he invites to an analysis of the leading principles and the direction of the narratives presented in the form of examples, instead of presenting a syllogistic logical procedure. To proceed by enthymemical argumentation may create an impression of nearness when used in front of an audience, but in a text such procedures often offer a Verfremdungseffekt, a feeling of estrangement in the reader. Biases and presuppositions are turned inside-out or pursued in their consequences.

In this respect, Wittgenstein's Nachlaß has a certain resemblance to the sceptical way in which Michel de Montaigne treated his own book of Essais, the genre he invented in order to be able to write. Into a copy of the first edition (1580) Montaigne inserted comments, additional examples, corrections, afterthoughts in the margins and on pasted-in pieces of paper. Montaigne continued his revisional work until his death in 1592. His personal copy of Essais, with all the alterations, was the basis for the posthumous sixth edition of 1595, just as he had ordained before his death. Montaigne stated that he wrote in order to understand himself. He was, in fact, one of the first moderns to problematize the knowledge of the self, and he did so by the continuous correction of his self in his own writings, in which he never ceased to question the validity of what he had heard, seen and read – what constituted the validity of his knowledge of his world.

We have for ever lost the voices of these authors' texts. But the traces that show how they reflected upon the voices and texts of their times are there, in the interpolations, the emendations and the comments they made. And it is exactly in the way such thinkers as these have gone about reworking their own language, that we may search for the answer to how subjectivity is constituted in such a way of writing. To put it otherwise, perhaps somewhat boldly but, as I hope, suggestively: it is actually possible to study how Wittgenstein, like Montaigne, develops a critical attitude towards the sources of the subject's own intentionality.

I haste to add that it is not the "author's workshop" of the romantic historicists that I suggest to look for in Wittgenstein's Nachlaß, nor is it the intellectual development of the historical individual Ludwig Wittgenstein. What I do think may be an awarding study, is the investigation of how this text material brings to light the process by which individuality is constituted, how a subject is objectivized in a carefully constructed manner and being manifested as a textual product.

In Philosophical Remarks, Wittgenstein remarks on the relation between subject and object:

§ 71: Visual space is called subjective only in the language of physical space. The essential thing is that the representation of visual space is the representation of an object and contains no suggestion of a subject.

Well, then: Aren't the texts at hand to be considered as visual space? Aren't many of the questions raised in Huitfeldt's paper questions concerning the visual representations of the text material? In textual criticism, anyhow, it is the visual appearance of the texts that first comes under scrutiny. More generally, the question of how readability and visuality is related in the understanding of texts, is a question of how the observance of sign patterns is turned into a critical examination of the uses of language.

Such a critical examination of the visual patterns of a text leads to reflections on the difference between representation and composition: A text is not the visual representation of a message in a similar way as letters are supposed to be visual representations of sounds. Writing implies the construction of interrelated patterns of visual representations and different levels of prescriptive rules for organizing complexities of written discourse. Among these rules are genre characteristics, which presuppose knowledge in the reader of how other, earlier or contemporary, texts are constructed. To study the complexity of a textual composition is, therefore, to critically examine prescriptive rules of how one should approach, describe, present and represent a subject or an argument. Such a study implies considerations on the ethos and pathos which is appropriate when dealing with a subject and proposing arguments within a certain genre. Such considerations constitute the starting point for the study of collective, that is, culturally determined, intentionality.

By now, I have already introduced my announced third area of questioning: the relation between orality and literacy, that is, of written discourse as opposed to spoken discourse.

Earlier, I quoted Wittgenstein's question: "If, when learning a language, speech, as it were, is connected up to action, can these connections possibly break down?" An answer to this question is that the connections between speech and action break down whenever writing represents discourse. Writing is another type of action than talking, reading is another type of action than listening. The writer is not necessarily conscious of everything that goes into his writing, but his construction of the text is nevertheless deliberate, he writes for some purpose – not necessarily for the purpose of conveying a message to someone (one might write diaries, for instance, to be able to get a clear picture of what has happened and in order to remember it later on). Anyhow, to put pen to paper or to open a new file on your harddisk is to go through motions that are learned and usually practiced for certain purposes.

Anyone who has had the experience of collaborating with another person on a manuscript (or even of writing a contribution to a collection of essays), knows that one adjusts to certain constraints of how to write on certain subjects – or to avoid writing in certain manners on certain subjects – as well as to an agreed division of labour. But this kind of adjustment – or the efforts made to avoid such adjustments – to constraints in the production of a text, is also present to the person writing in solitary confinement. Such literate – or "scriptural" – constraints are what makes reading learnable, but the knowledge of them is not necessarily sufficient to make a text understandable to a reader.

The reader may ignore what kinds of constraints the author was under during the time of the production of the text he is reading, or he may on the contrary be acutely aware of them. The reader may even read deliberately at cross-purposes, he may freely choose to search for other kinds of information in the text than the writer of the text intended a reader to look for. The situation of the reader, that he has such a possibility to choose among interpretational options and to weigh alternative ways of understanding the text, already shows that the act of reading cannot be symmetrical to the act of writing.

Reading is no simple decoding of encoded authorial messages, but implies constructive cognitive work on the part of the reader. The reader brings his experience of other texts and of his other practices to bear on what he reads. This is why many an old text never ceases to surprise readers. The act of reading may transform the way in which the reader looks at the constraints in his culture or in his participation in different fields of practice – different "language games". A text should, therefore, from the reader's viewpoint be regarded as a dynamic entity, where meanings in a well-constructed text may change over time according to such possibilities as are present to the reader.

Written and oral discourse have different organizational principles. This problem is systematically ignored by analytical philosophers discussing the function of speech acts. Accordingly, they tend to concentrate on single phrases or propositions and to furnish those with imagined contexts of oral communication, as if every kind of language use emerges in speech situations. Their "logocentrism" makes them unprepared or unwilling to discuss, for instance, the function of texts when read aloud or silently by language users, and it creates a blind spot in their analyses of communication in a literate society.

One very important feature in any text is that it sets up an alternative context to the speech situation when referred to or used by interlocutors in speech. Every phrase in a text is understood in its relation to other phrases in the text, and this relation is normally not just given by the linearity or the series of phrases following each other. The totality of the text is governed by aesthetic, rhetorical, grammatical and sometimes logical principles, determining the syntax of the phrases and thereby the semantics. Understanding of a text also entails knowledge of other texts of similar or different types, often a whole series of texts within a particular tradition. This is why texts commonly considered to be "classical" tend to change meaning over time, as interpreters have gained knowledge of other texts relevant to understand the genre or the intertextuality. The Bible, Homer, the fragments of the sophists, the works of the two Senecas or the works that were attributed to Duns Scotus are all examples of text complexes that are quite differently interpreted now than, say, two or three generations ago.

In oral communication, discourse is not arrested or fixed as in written discourse. Semantic determinants of oral language use are of other kinds, mostly social: conventions of how to behave towards people, power relations, the knowledge or the lack of knowledge that interlocutors have of each other, the actual historical situation, the corroborate purposes of the collective speech situation or the particular aims of individual participants. The difference in organization of written and spoken discourse is experienced as soon as we encounter a thought-figure like irony, where a meaning contrary to the explicitly expressed one is conveyed. In a speech situation, the understanding of ironical comments implies an acute awareness of the positions of the interlocutors, whereas textually formulated ironical comments cannot be appreciated without an understanding of how textual elements are related in the construction of a contrarywise way of referring to something.

When speech act theoreticians pick out propositions for closer scrutiny, they tend to treat spoken phrases as if these were cut-out text fragments thrown haphazardly into the world. One of the famous examples of John Searle is: "The cat is on the mat". Normally, we don't go around telling each other of cats' whereabouts. If we consider a normal situation of small-talk, it can be of greater consequence to pay heed to the way in which we often achieve an ironical effect by metaphorizing statements on the scheme of ostensive references: "As you know, Peter is a real tomcat. I'm rather worried about his health, because yesterday he sat on the mat all night!". This metaphor functions as a deliberately planned category mistake where characteristics of species are mixed, the effect being that Peter is classified – perhaps as admirable, perhaps as contemptible, anyhow as someone who distinguishes himself by his behaviour. In order to uphold a conversation, we do not reduce suchlike metaphors to their elementary referential properties, we accept the language game and jumps between classificatory strategies, and new clues may lead us on to new variations on a theme. In contrast, in a discussion of the metaphoricity of Baudelaire's sonnet poem Les chats, we are confined to the stratagems and structure of the poem when we want to elucidate what categorizations the text allows for. Searle's cat could have found a natural habitate in a nursery rhyme or in an elementary textbook for the first grade in primary school, in such cases we should consider how the rhyme is used to develop children's awareness of distinctive factors in language by playing with words.

In Les mots et les choses, Michel Foucault discussed how the early classifications of natural science, commerce and literature organized the world during the emergence of modernity. More generally, one could say that the impact of the text upon the world consists in the ordering, reordering and sometimes even desorganizing of our views of experience and information.

An obvious difference between oral and written language is that writing may be introduced into spoken discourse, whereas spoken discourse cannot be introduced into an already existing text without altering the text by writing and thereby creating a new text. The point is not just a banality; the classical scholar Eric A. Havelock has argued that the awareness of such a difference was the starting point for the ancient Greeks' awareness of history as well as for the necessity of logic. A text is an object with a history; there was something before the text was written, and something has happened since, there is one situation before reading the text an another situation afterwards.

To represent historical development by writing down a story is to fix the telling of events within a certain representational pattern, to discuss the interpretation of the story is to compare the validity of this pattern with the validity of conventional conceptions or formulas known to the participants in the discussion. Therefore, referring to representations of history implies considerations of causes and effects, not only as presented by a story written down, but also of any text: A list of objects classifies the objects listed as belonging to a type or types, a code of law sets some normative statements into the world, an edict orders people about, a story may beg questioning or provoke conflicting interpretations.

Much has been said and written during the last few decades on differences between oral, traditional societies and literate, modern societies. I should not want to impose such discussions upon the Wittgenstein Archives, only to suggest that it might be relevant to consider how the state of much of the text material in the Nachlaß in some respects resembles what one might find in a "manuscript culture", like monastries and universities of the late middle ages. The uses of manuscripts are of other kinds than the uses of print; manuscripts have a much more restricted circulation and copying them invariably creates more or less subtle changes both in their appearance and in their content. Before the coming of the printing press, manuscripts were mostly written to be read aloud, and contributed therefore to the ordering of oral culture within certain institutionalizations. We have maintained similar uses of manuscripts, like reading papers in a conference or circulating manuscripts for comments, or even the introduction of a new work into scholarly dialogue by the doctoral dissertation. In such instances, the texts constitute the agenda and they govern the individual and collective performances. Such uses of texts often contribute to a ritualized freezing of the speech situation, conferences seem to be a mode of existence suspended from the flow of everyday time.

But in the case of Wittgenstein's Nachlaß, the author hasn't allowed his texts to remain fixed entities in well-ordered series. The author seems to use his own texts as arms in a struggle between what he once thought and wrote and what he now wants to write and think. When Wittgenstein ordained that translations into English of his writings always should be accompanied by the German original when published – an ordainment not always respected –, one of the reasons could have been that the reading then would allow for a comparison between language games, stylistic nuances and semantic possibilities in the two languages.

In the Scandinavian countries ever since the great pedagogical reforms provoked by Grundtvig in the first half of the last century, we are wont to praise "the living word", the liveliness of spoken discourse, as the primary source of insight and understanding. But there is a distinct possibility that speech may make it more difficult for us to find truth, because we are lead to accept intrinsic conventions and norms that rule oral communication, and the costs of questioning conventional usage might be so great that we often refrain from such activities, unconsciously or consciously. To study Wittgenstein's way of writing could challenge the presuppositions of such logocentric traditions. By returning to his own texts in order to alter them, Wittgenstein arrests conventions and struggles with those thought-patterns that he wants to free himself and others of. Thereby he shows how writing gives possibilities to question received views on how language functions.

Thus, and finally, my fourth proposed area of questioning comes up in the guise of a conclusion: the possibility of a critical hermeneutics.

Textual criticism implies critical examination not only of the form of the texts that are present as objects for study, but often also of beliefs and notions that have guided former or contemporary interpreters. Literary criticism implies evaluations of the validity of interpretative strategies as well as particular interpretations of texts. Rhetorical criticism may be said to imply a critical study of the circumstances under which persuasion is possible, and this is a tradition that might be brought to bear on the question of how the uses of texts contribute to the construction and identification of collectively recognizable intentionality.

All these traditions within textual studies have bearings on the way one treats epistemological problems in the uses of texts. Therefore, critical philosophical analyses of texts should pay heed to the lessons to be learnt from philological investigations, and philological examinations of textual material should pay heed to the philosophical problems involved.

To take into account the interrelatedness between philosophy and philology is altogether another endeavour than to simply promote collaboration between experts within different disciplines which boast their own particularities in problems and problem solving. It means that we should pose other problems than those commonly accepted within the disciplinary matrixes, that is, the paradigms. Analysing the construction of subjectivity and intentionality in texts implies considerations on how language unwinds and is understood within time.

In order to approach the question of how the understanding of language itself is culturally bound, these bindings or determinants should be critically studied as productive entities, patterns being introduced into and guiding the situations of writing, reading and talking. Such an investigation of how texts function as dynamic entities constitutes a critical hermeneutics, taking into account the cultural historicity of concept complexes.

My suggestion is that the designation "critical hermeneutics" could be taken to refer to the study of interpretational matrixes within particular cultural settings. Perhaps my use of the notion of "interpretational matrix" is in need of some elucidation. It could be considered as analogous to Kuhns "disciplinary matrix" – commonly accepted guidelines, ways of conducting research and teaching, institutionalizations that ensure the production of knowledge within the reproduction of a common understanding of how a discipline functions.

An interpretational matrix is, then, the ensemble of available patterns of interpretations, the stratagems for understanding how to acquire knowledge, the procedures for validating interpretational possibilities.

In the comparisons between interpretational clues of available texts and the clues of interpreting past experience, reside possibilities for the critical examination of the construction of personal as well as collective identity.

Willard Van Orman Quine once said, in his much-quoted work From a Logical Point of View:

Identity is a popular source of philosophical perplexity. Undergoing change as I do, how can I be said to continue to be myself? Considering that a complete replacement of my material substance takes place every few years, how can I be said to continue to be I for more than such a period at best?

Quine suggested that we ought to consider how abstract entities gain their hold upon our imaginations, so that we too readily commit ourselves to the use of general terms in descriptions of identity, a procedure which makes us confuse general terms with singular ones. He proposes a pragmatic view on what he calls our "conceptual scheme", we can change and improve our eclectic heritage bit by bit while continuing to depend upon it for support – like Neurath's philosopher, who is compared to a mariner who must rebuild his ship on the open sea. Quine flatly declares that we cannot detach ourselves from our conceptual scheme "and compare it objectively with an unconceptualized reality".

Quite obviously we cannot un-conceptualize the experiences which constitute our relations to the social and material world, we are in need of concepts and symbols to be able to think. But what we can do, is to compare our conceptual schemes, or rather our interpretations of ourselves, with those interpretational matrixes offered by the stories, the examples and the classifications presented to us by the texts of our culture. Such comparisons between language games allow us to break out of the conceptual boundaries of the particular language games we exist within, and to take on responsibility for intentions that may go against the intentionality commonly accepted within the practices we partake in. This analogous mode of reflection constitutes, I do believe, an ethical significance in Wittgenstein's encouragement to compare examples. It makes it possible to analyze and even criticize assumptions integrated into our tacit knowledge.

Quine's approach to the understanding of how we pragmatically insist upon being ourselves by being aware of how we change, seems to me to suffer from his own insistence upon the methods for identifying objects conceptually. What of our recognition of the ambiance of a refreshing friendly discussion, or of the enmity of opponents in budgetary discussions in a faculty board, what about the identification of the point of a story, the usefulness of an example for an argument, the stylistic appropriateness of elaborations in a speech, the pleasure of being surprised by the turns of a poem read silently or aloud, the annoyance of listening to a bad singer? These kinds of identifications cannot be pinned down, neither by general nor by singular concepts. They do, however, constitute ways of learning and ways of judging ourselves and others. And the interpretations such reading of identifications entail, may be compared to alternative or similar patterns of interpretations in texts. By such an approach to texts, we establish a kind of "participational objectivation" in our identification of our own selves and of the possibilities of intentions.

Like the aforementioned scientists in their laboratories or seminars, we tend in everyday life mostly to comfort ourselves by reascertaining our experiences. In some respects, Quine's improved conceptual scheme and the ship of Neurath's mariner are variations on this theme. One would suppose that Neurath's mariner would have difficulties in rebuilding, say, a schooner into a destroyer (and where is Neurath's lumberer?). Quine's use of Neurath's metaphor invites us to reassure ourselves that the more we change, the more we remain the same, as much unaffected by passing pirates and supply ships as by visits to foreign shores. Another well-known metaphor, that of the "encounter" between reader and text, is often interpreted somewhat similarly, in that it is taken to mean that the reader assimilates into his awareness of his own inner being whatever he enounters when reading. But the trouble is that texts often prove to be very uncomfortable to our self-consciousness. At least, the history of reading habits tells us that a self-comforting smugness is no necessary result of reading. On the contrary, texts have been known to alter readers' experiences of themselves quite thoroughly.

Thus, a critical hermeneutics implies efforts to understand how writing and reading might alter self-consciousness and provoke change in culturally bound contexts. The study of Wittgenstein's Nachlaß under such a perspective could tell us something of how consciousness of the possibility to compare language games makes it possible to break some of the boundaries set up for us in language games.

Refbacks

• There are currently no refbacks.