© Jan Christoph Meister. All rights reserved. Version
03.09.2003. Contact: mail@jcmeister.de
Jan Christoph Meister
Universität
Narratology as Discipline: A Case for
Conceptual Fundamentalism[1]
“The king died, and then the queen
died, and then she gave birth.” – With a bit of logical laissez-faire one can comfortably proceed from a minimal to a never
ending story. Narratology, it seems, has
mastered the trick: from paradigmatic and rock-hard ‘science of literature’ to
best-hated victim of postmodernist critique and deconstruction, from sterile
logo- and phallocentric villain to revered disciplinary ‘great-grandfather’
(Nünning) or rather, great-grandmother whose offspring resembles a family of
old-testamental dimension. In a comprehensive survey preceding the one
presented in the current volume Ansgar Nünning already lists no fewer than 35 ‘new
narratologies’[2] in six categories, while Monika
Fludernik identifies four ‘new directions’, one of which – a case of
retro-autogamy? - includes ‘traditional narratology’.[3] Seen in this light David Herman's
suggestion is perhaps the most plausible of them all: he advocates putting the
term narratology into the plural anyhow and using it as “interchangeable with
narrative studies”.[4]
The straightforward progression from historical reconstruction to programmatic outlook concerning narratology's development and proliferation may or may not lead to paradoxical consequences: a rather common characteristic of such attempts is the use of a tentative and fuzzy descriptive terminology. What, for example, is the 'new' narratology's or narratologies' proper methodological label? One recent survey refers to it alternately (and within a mere 7 pages) as a ‘school of thought’, a ‘scientific praxis’, a ‘discipline’, a provider of ‘terminological and descriptive tools’, then again as a ‘discipline’. It raises the question whether “narratology is, indeed, a critical school like deconstruction” only to state five lines further that “if at all, narratology is merely a subdiscipline of structuralism”. Yet the concluding paragraph offers comfort by declaring narratology “a flourishing discipline” once more.[5] Given such uncertainty about so-called 'classical' narratology’s identity one can hardly expect its progeny to do any better. Thus Nünning’s investigation into the “rise, fall and renaissance” of narratology sets out similarly by quoting Todorov’s definition of narratology as the ‘science of narrative’ and by referring to narratology as ‘a flourishing discipline’ (this time pointing specifically to that of the 1960s). However, the blossoms produced ─ the plentiful offspring of 35 ‘new narratologies’–for some reason only qualify as ‘sub disciplines’. Three pages later they are demoted to mere ‘approaches’, and eventually they resurface in an even more modest form, namely as (interdisciplinary) ‘projects’.[6] So what is narratology - approach, praxis, project, school, sub-discipline, discipline, science? And/or which narratology is what?
The merits of these and similar attempts at reviewing narratology’s history and outlining its potential future development need not be re-emphasised. To present a new taxonomy for an entire field of scholarly endeavour is to take an equally bold and necessary step. However, the frequent shift in categorisation which we observe in these taxonomies clearly demonstrates that merely descriptive surveys experience considerable difficulties in coming to grips with questions of principle, and particularly with the problem of defining narratology's methodological identity. The question thus remains: "What is Narratology?"
In the following I will resort to a more systematic and fundamentalist approach, hoping that it might complement the more empirically orientated surveys presented by Ansgar Nünning and others. The main points that I will attempt to argue are:
· One: the fundamental definition of narratology is to be that of a discipline, that is, a system of scientific practices for research into the conditions of possibility of an object domain called ‘narrative’.
· Two: claiming disciplinary status does not entitle narratology to recede into ‘splendid isolation’. As a Human Science narratology has to remain ‘anschlussfähig’, that is, it has to provide conceptual interfaces to other systems of practice of diverse methodological status (whose own status might indeed range from approach to sub-discipline, or even to that of a fully-developed discipline).
· Three: it is therefore particularly important to define the methodological ‘terms of trade’ between narratology and its potential partners in order to ensure that the integrity of the procedures and terminology that define narratology’s disciplinary identity be maintained.
On a more general level I am taking
up a suggestion by Meir Sternberg who only recently ─ and quite
categorically ─ stated that the “disciplinary foundations” of narratology
have not even been laid. [7]
In other words: whatever praise our historical reconstruction of the
‘golden days’ of Structuralism wish to sing, and irrespective of the great
plans we might hatch for the hopeful narratological offspring, we must go back
to the fundamentals. Sternberg believes that the following two questions need
to be clarified: 1) “What is narrative?” and 2) “What becomes in it of the
components shared with other genres?” In
other words, he wishes to define narratology as a discipline by explicating its object of study in
terms of its generic features.
Though focusing on the same question of disciplinary status my own approach is different in that it does not concentrate on the generic features of the object domain, but on the procedures that can guarantee the narratological system of practice’s integrity. More particularly, I will try to sketch what I believe to be a rational modus operandi for formulating narratology’s fundamental concepts under the methodological constraints of disciplinarity.
The first question to be addressed is thus that of the theoretical definition of ‘discipline’ itself. Martin Guntau and Hubert Laitko, two of Germany’s leading historians of science, have taken as their starting point our intuitive contemporary understanding of disciplines as object orientated system of scientific practices.[8] In this regard we must first distinguish between the pragmatic and social circumstances that govern any discipline’s eventual institutionalisation, and the more fundamental theoretical definition concerned with the explication of the specific qualities that set a scientific (in the sense of: scholarly or ‚wissenschaftlich‘, as opposed to the narrower Anglo-American understanding of ‘scientific’ as ‚naturwissenschaftlich‘) object orientated system of practice apart from non-scientific practices. It is crucial to realise that the notions of ‚object‘ and ‚practice‘ refer to different things in their scientific and non-scientific understanding respectively:
· non-scientific systems of practice are based on a notion of ‘object’ which is understood to signify a specific class of empirical objects, whereas scientific systems of practice are based on a notion of ‘object’ which is understood to signify the system of properties and relations which are shared by all empirical objects in a class.
· in a non-scientific context ‘practice’ is conceived of as engaging in a ‘doing’ with or on the empirical object. The purpose of this doing is to reach a final pragmatic goal, whereas in a scientific context ‘practice’ is conceived of as engaging in investigations into the generalised system of properties and relations. Assuming that it is theoretically impossible to ever reach the stage where all possible aspects of such a system have been explored the scientific practice is in theory ever-perpetuating (though in reality of course limited by pragmatic and social constraints).
Clearly, more than one such scientific
system of practice may investigate an
identical class of empirical objects while conceptualising them as different scientific objects, i.e., different systems of properties and relations.
Take for example the corpus of Russian fairy tales studied by Propp which was,
however, clearly not the exclusive domain of Formalist study, but also of
interest to ethnological research projects.
This observation immediately raises the question of how to tell the difference between competing
conceptualisations of a particular object domain, and thus the difference
between the various disciplines that
engage in its study. The fundamental
difference lies, according to Guntau and Laitko, in the procedures whereby the
scientific object is identified, and in the terminology whereby it is
subsequently described. Both ─ procedures and terminology ─ are
highly discipline specific. Secondary conceptualisations that render complex
models and unique theories of the object domain might be seen as the more obvious
manifestations of disciplinary variance. But the root cause of their divergence
is to be found in the respective scientific practices' unique and most basic
methodological rules determining how to handle the object under investigation,
and how to name its constituent parts and features.
Ideally speaking the integrity of a given discipline is thus the function of a consensus on procedures and terminology shared by those who subscribe to its system of practice. But disciplines would become stale and static if they were to make compulsory the use of a particular terminology: scientific progress is to a large degree dependent on constantly revisiting, refining and enriching one’s terminology. If terminology can and indeed must change over time, then the bulk of the responsibility for ensuring a discipline’s identity rests on the agreement on rule-based procedures. The essence of these rules is to guarantee that in following a given procedure, we will not just elaborate further on the scientific object as a system of properties and relations, but will also be able to relate these observations back to identical empirical objects intersubjectively, and at any point in time. This is the necessary disciplinary pre-condition for building ever more complex models and theories.[9]
We can now address the question of how to construct a terminology. Let us assume that there is a class of empirical objects called ‚narratives‘. Our task as narratologists would then be to elaborate at least one, if not an entire body of narratological theories each of which at the very minimum must fulfil the conditions that define a valid scientific description of the class as a system of properties and relations. An optimal valid description will be one that reflects on all the conditions of possibility of any member in the class labelled ‚narratives‘. A sub-optimal description may ignore some of these conditions; however, the more such conditions it ignores, the greater the risk that we will not be able to reach an agreement on how to bind it back to identical empirical objects ─ particularly when we try to evaluate the adequacy of more complex theories and models that were built on the basis of these sub-optimal descriptions. These descriptions must therefore be traced back to the fundamental level where we assign specific terms to specific phenomenological entities ─ features, qualities etc. – observed in the object domain.
To make things less abstract let us now concentrate on one such feature which most ─ but, as the examples of Fludernik or Sternberg show, not necessarily all ─ of us intuitively deem to be a necessary condition of possibility in every member of the ‚narratives‘-class. This feature is that of symbolic representation of a structural real-world phenomenon called an ‚event‘. If we are right in our intuition then we can identify two major procedural requirements to be met by our undertaking. One: we need at least one element in our terminology which will enable us to describe the ‚event‘-feature in any empirical ‚narrative‘. Two: such a term must fulfil more than just the referential function of pointing from ‚within‘ narratology to an aspect of any of the empirical objects falling under the class-label ‚narrative‘. The term must also function as a concept, meaning that it must be possible to integrate it into every system of properties and relations that successfully models the entire class, that is, that represents a scientific vs. the original empirical object. This system will most probably contain a good deal of other such concepts (assuming that members of the ‚narrative‘-class have more than just the ‚event‘-property in common).
Taking these factors into account we can now attempt to visualise the terminology building process for the concept of ‘event’ in the form of a ‚workflow‘ graph. The graph tries to show how this particular concept is arrived at by way of abstracting from a specific constellation of textual features, in this case, the juxtaposition of two propositions with a shared argument and opposing predicates:
Concept generation workflow 1: Abstracting a fundamental concept from empirical data
Admittedly, every fundamental
descriptive term will at closer scrutiny turn out to be rooted in an implicit
frame of reference; its 'fundamentalist' status thus being of a relative nature
only. However, in terms of the workflow
it is important to note that at this point we are still exclusively engaged in
an inductive bottom-up process: we consider the formation of our narratological
descriptive terminology only, disregarding the subsequent integration of
elementary concepts into more complex clusters that begin to interact as a
proper system of properties and
relations. In methodological terms
there is a great deal of merit in turning a blind eye at the various elementary
concepts' frames of reference at this early stage. If each and every of our
elementary descriptive terms ─ say, not just that for the
‚event‘-feature, but also those for the ‚mediation‘- and the ‚anthropomorphic‘-features
─ were taken from the same
theoretical frame of reference we would run the risk of already enforcing such strong conceptual
homogeneity at base level that our own narratological theory yet to be
constructed would eventually result in little more than a mirror image of the
theoretical frame of reference that influenced our choice of terminology.
Building a descriptive taxonomy on the conceptual level is one thing, but building a relevant taxonomy may turn out to be another. Ludwig Wittgenstein, I believe, pointed to this when he postulated in the Logische Untersuchungen: „All explanation must disappear, and mere description must take its place. And this description receives its light, i.e. its purpose from the philosophical problems.” [10] In other words: a description for description’s sake is not the solution, but one must be even more aware of rendering unwarranted explanations of empirical phenomena. The phenomenon as such has no inherent problematic, and hence there is nothing to explain and its description will indeed suffice. But there are of course questions that come to our attention and call for an explanation as soon as an aspect of the empirical object is accentuated by the ‚light‘ of an external ‚philosophical problem‘.
For Wittgenstein philosophical problems were rooted in an inadequate use or understanding of language. This analytical notion of a ‚philosophical problem‘ might be too restrictive in our case. The ‚philosophical problem‘ that sheds a light on, and thereby renders purposeful the narratological description of its object domain must be understood in a wider sense. It is represented by externally motivated questions that concern aspects which are not those of singular features of a member of the ‚narratives‘-class, in other words, idiosyncratic features represented by a singular entry in our terminology. True ‚philosophical problems‘ and questions must rather deal with effects occurring (and re-occurring) on the system level. They are effects whose description necessitates complex terms, if not entire models. ‚Event‘, for example, for as long as we pre-define this entry in our terminology in a strict bottom-up process as indicated in the workflow scheme, is a fundamental concept denoting an elementary feature of a given member of the object class ‚narrative‘. ‚Action‘, by contrast, refers to a structural compound made up of singular ‚events‘ that has been generated according to certain systematic rules ─ thus ‚action‘ already is a complex concept. It is hard to conceive of a genuine ‚philosophical‘ or, for lack of a better word, ‚hermeneutic‘ problem attaching to a phenomenon on the level of isolated ‚events‘, whereas our capability of identifying such problems on the systematic level of ‚action‘ seems limitless.
What does this mean in terms of the proposed terminological ‚workflow‘ model? Complex concepts cannot be generated in a mechanistic fashion by simply grouping together a subset of fundamental terms between two brackets like, for example, ‚event‘ and ‚character‘. Complex concepts like ‚action‘ will always inherit an additional element, which in the case of the particular concept of ‚action‘ is an import from a second theoretical frame of reference. We must therefore extend our workflow model accordingly and make it permeable so that conceptual enrichment can take place. In the case of 'action' we might for example choose to add an interpreting predicate to this complex concept that explicates its moral value ─ an operation necessitating access to a 'moral' frame of reference:
Concept generation workflow 2: Building of complex concept by explicating a frame of reference
‚Moral philosophy‘ is just one of
many potential explanatory frames of reference by which we can generate a
particular complex concept of ‚action‘ that combines the fundamental
descriptive concepts ‚event‘ and character‘ in the true sense of a partial sub-system of properties and relations
within the global system that models the scientific object ‚narratives‘. The combinatory operation taking place at
this level is mainly a deductive top-down process which seeks to bind together
particular fundamental concepts according to a theory, model or hypothesis
which has been formulated outside of narratology proper.
What about the other option: doing away with the constraints of a distinct terminological frame of reference restricted to the generation of fundamental terms, and short-circuiting it with the systematic frame of reference right from the outset? Wouldn’t this yield far more powerful narratological concepts, concepts that offer such high level of systematic connectivity that we could come up with far more ambitious narratological systems of properties and relations modelling the scientific object ‚narrative‘?
This is the alternative which I understand to be advocated by some of the more recent attempts at re-defining narratology on a whole. It is an approach that amounts to treating narratology with what I would like to label a ‚methodological fabric softener‘ which will indeed affect the conceptual fabric at base level. ‚Natural narratology‘ in the variant proposed by Monika Fludernik, for example, eliminates the notion of ‚event‘ from narratology’s set of elementary terms and concepts, and proposes to replace it with the high-level concept of the ‘natural’ as “that which is anchored in humans’ everyday experience”.[11] One of this concept’s theoretical frames of reference is Cognitive Theory - which in itself is a fully developed discipline. The consequences of Fludernik’s proposal are far from modest: Not only is the consensus on what forms narratology’s scientific object immediately put into question (we could simply argue that the old system of properties and relations has to be replaced by a new one), but it has become completely impossible to relate the scientific object in whatever form back to identical empirical objects unless we all agree to throw the concept of ‘event’ out of the window.
But let us not
focus on what arguably may constitute an equally bold and extreme case. The
problem which I would like to bring to attention arises much earlier. For
example, the Narratological Research Group at
This is not to say that there is no need for such a complex concept. There might very well be. The problem however lies in the fact that we must either accept that a fundamental and a complex narratological concept share the same name:
Concept generation workflow 3: Conflict between fundamental and complex definitions of a concept
This is bound to lead to confusion,
but could be remedied by way of a new terminological convention. Rather more problematic is the second case
where the original fundamental concept of ‚event‘ is replaced by the new
emphatic one altogether:
Concept generation workflow 4: Elimination of fundamental concept, resulting in a 'baseless' complex concept
Terminological and conceptual
erosions like these will ultimately lead to the demise of narratology as a
discipline. To re-declare elementary concepts as complex
ones is to invite non-agreement and fundamental dissent concerning the actual empirical objects which are represented,
in the form of a system of properties and
relations, as narratology’s scientific
objects. Of course, this entire argument hinges on my premise that narratology
is, or ought to be, conceived of as a discipline.
What are the conclusions to be drawn
from this examination of the fundamental concept generating operations
necessary for the formation of a discipline with a consistent and transparent
scientific terminology?
One: we should learn to appreciate (instead of defending or excusing half-heartedly) the formal and context-free nature of structuralist narratology’s fundamental concepts. It is exactly the level of formalisation and abstraction which has made the ‚narratological toolkit‘ so accessible to related disciplines and provided – and still provides ─ us with interfaces to complex ‘philosophical questions’.
Two: Toolkits are sacrosanct. Smuggling a cognitivist, hermeneutic, psychoanalytical or otherwise -isted, -ized or -alled high-level flexible conceptual spanner into our fundamental ‘toolbox’ amounts to a breach of the procedural rules that define any discipline – and not just narratology ─ on the systematic and functional level. Let us take the ‘toolkit’-metaphor seriously: trained mechanics don’t use shifting spanners; however, they can of course construct and manipulate the most intricate machines by putting together various components. Likewise, conceptual amalgamation and recombination in a scientific discipline can only take place at the level of complex concepts and systems. It would therefore make good sense to identify which of narratology's concepts are elementary non-negotiables, and which aren't.
Three: If narratology is indeed a discipline, then there can by definition be one narratology only. The branching out of disciplines – as it often occurs in the natural sciences: take for example biology which has differentiated into molecular and micro-biology ─ is normally a result of the pragmatic need to break up an over complex domain of empirical objects into manageable sub-sets which then constitute separate scientific objects. Most of the ‘new narratologies’, however, aim at treating the old object domain with a new amalgamation of concepts and procedures. In other words, they promote narratology’s differentiation or re-orientation on methodological (if not ideological), and not on empirical or pragmatic grounds. Given this the flags under which they sail might in fact be part of the problem. Instead of labelling them with compound terms that use ‘narratology’ as a generic term one should perhaps decide to use it as an attribute that refers to the new practice’s dominant (if indeed it is dominant!) methodological heritage. Thus it might make better sense to talk of, say, ‘Narratological Cognitivism’ than of a ‘Cognitivist Narratology’. [12]
Four: None of the above advocates turning narratology into a ‘closed (work)shop’ for hardened structuralists again, or intends to disallow the critique of its fundamental concepts. This critique is absolutely vital, and, as Meir Sternberg has rightly pointed out, many of our most basic definitions may indeed be far from satisfactory. But this is exactly why one should not get carried away with the multitude of applications to which the narratological methodology is now being treated. Narratological fundamentalism, if properly understood, might well help us to discover or create even more powerful conceptual interfaces with other fields of study.
Finally, narratological fundamentalism is by no means an over modest practice; rather on the contrary. As Heidegger put it in ‘Sein und Zeit’: “Das Niveau einer Wissenschaft bestimmt sich daraus, wie weit sie einer Krisis ihrer Grundbegriffe fähig ist.”─“The standard of a science is dependent on the extent to which it is capable of a crisis of its fundamental terms”.[13] Seen in this light a systematic re-examination of the basic concepts of narrative theory and narratology might not only result in a better understanding of how 'narratological' the new narratologies in fact are: it might also help to clarify whether narratology is indeed a discipline in its own right ─ and if not, what else it can be, or should.