One pressing problem seems, then, to be the need for the development of more sophisticated filtration methods and disambiguation devices for recovering items from large quantities of knowledge and information. Digital communication and media technologies already provide potential gateways to enormous amounts of on-line information, and as human beings we have only limited capabilities of sorting and integrating this into our individual and collective life-worlds. Facilitated informed choice is rapidly becoming a necessity. We still know far too little about the dynamics of the evolution of human knowledge, and about how such knowledge is socio-semiotically constituted and filtered in dynamic open systems like language and culture [see also Coppock 1995[c]].
Distributed hypermedia systems, of which the World Wide Web is the best available example may also be considered dynamic open systems, and this necessitates a new understanding of the concept of the "node" mentioned previously in the more traditional conceptualisation of the hypertext structure. From the idea of a node as a more or less static set of semantically organised knowledge representations constituted as texts, we must begin to understand the node as a dynamic gateway or a window onto social fields of culturally instantiated and constituted understandings and meanings which are constantly changing, and which are being constantly revised and updated. This is reflected concretely at the level of the user-interface by the dynamic material links between documents distributed throughout the World Wide Web, often known as "Home Pages". The links made accessible from any given Home Page reflect and index in various ways the interests of the various people and the socioculturally constituted institutions "behind" the page in question. They also reflect the process of meaning-making as it occurs. It is not uncommon, for instance, while "browsing" such pages to find that previous links to other pages of related interest have been removed or updated from week to week, or even from day to day. A common icon to be found in World Wide Web pages is the little "road sign" with a man in a hard hat shovelling dirt symbolising "work in progress". Textual messages such as "This page is still under development", and "Watch this space for more information!" abound. Each page is thus a material representation, instantiated and construed through the use of an increasingly wide range of semiotic systems and codes, of a social field belonging to some area of the wider dynamic open systems of culture and science. Later I shall briefly mention other, even more dynamic material representations of such social fields, namely distributed text-based multi-user virtual environments, or MOO's, where the material instantiation and construal of meaning through interaction in written language may be studied in real time. First, however, I would like to discuss some issues related to the disambiguation of meaning as a social process.
In a recent report from SRI: "Interpretation as abduction" [Hobbs, Stickel, Appelt and Martin 1990], Jerry Hobbs and his co-workers discuss some problems related to interpreting texts. In their perspective, interpretation of the text is "a minimal explanation of why the text would be true". One must "prove the logical form of the text from what is already mutually known, allowing for coercions, merging redundancies where possible and making assumptions where necessary." [Hobbs &al. 1990, p.1] This group denotes the problems they are working with as "locally pragmatic"; involving "reference resolution, interpretation of compound nominals, resolution of syntactic ambiguity and metonymy, and schema recognition". Their proposed solution is by means of what is called "abductive inference", inspired by philosopher Charles Sanders Peirce's concept of abduction (or retroductive hypothesis). Hobbs & al. define abductive inference as "inference to the best explanation", or alternatively, viewing the process of "interpreting sentences in discourse [...] as the process of providing the best explanation of why the sentences would be true". In their conclusion, they write that:
"Interpretation in general may be viewed as abduction. When we look out of the window and see a tree waving back and forth, we normally assume the wind is blowing. There may be other reasons for the tree's motion; for example, someone below window level may be shaking it. But most of the time the most economical explanation coherent with what we know will be that the wind is blowing. This is an abductive explanation."
This definition of abduction is however considerably more simplistic than Peirce's own formulations of abduction, one of which states that:
"Presumption, or more precisely, abduction [...] furnishes the reasoner with the problematic theory which induction verifies. Upon finding himself confronted with a phenomenon unlike which he would have expected under the circumstances, he looks over its features and notices some remarkable character or relation among them, which he at once recognises as being characteristic of some conception with which his mind is already stored, so that a theory is suggested which would explain [that is render necessary] that which is surprising in the phenomenon." [CP 2.776]
and elsewhere:
" The surprising fact, C, is observed;
But if A were true, C would be a matter of course,
Hence there is reason to believe that A is true." [CP 5.189]
For Peirce the "surprising fact" does not correlate with previous knowledge, and abduction is a necessary device to focus attention on the relations between aspects of features of the phenomenon which may seem characteristic of already existing conceptions. Once an exploratory hypothesis is formed (through abduction), the investigator then has to use induction and deduction to formulate and test empirical investigations which might establish the practical consequences of the hypothesis being true. Abduction in the true peircean sense is thus more speculative ("presumptive") and innovative than the mere "filling in" of non-perceived causal relations implied by the example of the tree given by Hobbs & al above.
According to Malinowski one must study human interaction in a context of situation in order to arrive at any deeper understanding of how cultural meanings are developed and constituted through language. As he points out:
"...it should be clear at once that the conception of meaning as contained in an utterance is false and futile. A statement, spoken in real life is never detached from the situation in which it has been uttered. For each verbal statement by a human being has the aim and function of expressing some thought or feeling actual at that moment and in that situation, and necessary for some reason or other to be made know to another person or persons - in order either to serve purposes of common action, or to establish ties of purely social communion, or else to deliver the speaker of violent feelings or passions. Without the imperative stimulus of the moment, there can be no spoken statement. In each case, therefore, utterance and situation are bound up inextricably with one another and the context of situation is indispensable for the understanding of the words." [Malinowski 1947, p 307]
In order to disambiguate situated complex meanings much more contextualisation of each utterance (i.e. accounting for the situation of the utterance) obviously seems necessary. There is also the issue of the interactions of the interpreter with the environment to take account of. Since Hobbs & al's method seems grounded in what Volos&inov has referred to as an abstract objectivist view of language [Volos&inov 1973], and the idea that it is possible to disambiguate meanings merely on the basis of "language internal" phenomena like reference resolution, compound nominal interpretation, syntactic ambiguity, metonymy resolution and scheme recognition etc., this is obviously problematic. Though they go on to outline further work with local pragmatics problems (i.e. lexical ambiguity, metaphor interpretation and resolution of quantifier scope ambiguities), recognition of discourse structure and the relation between the utterance and the speaker's plan and the drawing of quantity and similar implicatures, none of these approaches will be able to address the issue of how meanings are attributed by interpreters to situated utterances, for instance on the basis of on-going estimations of the relative importance of the utterance in relation to other on-going socio-semiotic processes of meaning-making occurring more or less simultaneously.
In order to solve disambiguation problems in language use I belive we need to move beyond an abstract, logically based semantics and pragmatics and into studies of what Umberto Eco has referred to as principles of "decimation" [see Coppock 1995[c]]. By decimation means the ways in which experts reduce the number of search choices they have to make in large amounts of information through knowledge of the norms, history and development of the field and the indexing to high quality information provided by established cultural norms and institutions. This means we have to find out more about how meaning is socially and culturally instantiated, actualized and construed, and how meanings are disambiguated by experienced researchers in specific contexts, as they often must be in new interdisciplinary fields of science where people with rather different intellectual, experiential and cultural backgrounds collaborate to reach common scientific goals.
[NEXT]