Disguised Encyclopedias? (flash movie)

[15] Umberto Eco's theories can help us understand the relationship between encyclopedias and literacy and the problems surrounding the construction of literacy curricula, especially E. D. Hirsch's dictionaries of cultural literacy. Eco's discussion of dictionaries and encyclopedias is highly technical, involving abstract notions of how an ideal dictionary might be structured. Beginning the chapter in Semiotics and the Philosophy of Language concerning these issues with the question, "Is a definition an interpretation?" -- Eco demonstrates that existing theoretical models for constructing definitions are untenable on several counts. Attempts to construct models for dictionaries based on the notion that a dictionary should store a finite number of bits of information about a particular lexical item and inventory a finite list of entries fail.

[16] Dictionaries theoretically rely on the idea of a list of semantic primitives (the simplest concepts which can be identified, e.g. human is simpler to identify than mammal). This list is devised in order to conceive of a dictionary-like competence free of any commitment to world knowledge; i.e., human is only identifiable as a result of experience. In other words, the dictionary attempts to make a list of primitives which can be understood without world experience. But, if one believes primitives are rooted in world knowledge, then dictionary competence is dependent on world knowledge.

[17] The Porphyrian tree is the model upon which dictionary definitions are built. The third century Phoenician, Porphyry, elaborated a theory of division based on Aristotle's Categories. While only suggested verbally in Porphyry's Isagoge, medieval tradition built the idea of a tree into visual representations tied to logical analysis. (59) Bacon's tree begins here. This analysis forms the basis for the construction of dictionaries with a finite list of entries and bits of information about those entries. Eco's claim is that the Prophyrian tree upon which such models are built ultimately yields not a finite list, but an infinite list, because the semantic primitives upon which a finite list are built remain rooted in world knowledge. His conclusion: it is impossible to construct a dictionary which is free of a commitment to world knowledge. Attempts to limit dictionaries to a list must fail.

[18] Definitions which take the Porphyrian tree as a model rely on dualistic division of the qualities of any item. Thus, man comes to be defined by the division of the corporeal (body) into animate (living being) or inanimate (mineral); living being into sensitive (animal) or insensitive (vegetal); sensitive into rational into mortal (man) or immortal (God). Each level of division supplies differentia. Differentia are qualities; differentia are expressed by adjectives. Differentia belong to infinite sets, not finite sets. The Porphyrian tree appears to be finite and ordered, but it is, in fact, infinite because the number of differentia needed to distinguish any item from any other is unknowable. Each node on the tree requires that we infer other differentia which are not named. Bacon's tree and then Diderot's encyclopedia grew according to this logic. For example, sensitive (animal) implies a contextual knowledge of the category animal that has experienced animals as sensitive versus plants as insensitive (a debatable point that also reveals the ideology of local knowledges). This contextual world knowledge necessarily draws on numerous associative networks. Eco makes clear that such networks are a priori infinite. Recent artificial intelligence research has attempted a solution to this problem by constructing semantic models that draw on world knowledge and by inventing the notion of "frames" and "scripts" that enable interpretation based on context. Computers are, of course, trapped in the binary logic of the Porphyrian tree. It seems unlikely that they will be able to break free of it as long as they remain binary “thinkers.”

[19] Eco continues his analysis by emphasizing that a dictionary attempts to be highly ordered, to include in its definitions the minimal needed to differentiate between signifiers. In order to hold together, a definition must ultimately rely on concrete and finite differentia. Paradoxically, a definition must explode into a multitude of differentia because the logical exclusion of entries and bits of information about those entries fails in the real world of semiosis. The necessary result is the illogical exclusion of differentia. So Bacon's tree grows through systematic exclusion, not just at the level of the limits of the primary list of definitions, that is at the level of what gets chosen for the list, but at the level of the differentia needed to sort one entry from another. The excluded bits point towards the world knowledge which is assumed to be unnecessary for interpretation. The result is that we are forced to infer the essential differences between entries based on our world knowledge; i.e., we must use world knowledge to interpret dictionary entries, but the dictionary represses our consciousness of this.

The tree. . .blows up in a dust of differentiae, in a turmoil of infinite accidents, in a nonhierarchical network of qualia. The dictionary is dissolved into a potentially unordered and unrestricted galaxy of pieces of world knowledge. The dictionary thus becomes an encyclopedia, because it was in fact a disguised encyclopedia. (68)

[20] But what are encyclopedias, and what kind of literacy do they require and enforce? Eco tells us that since dictionaries are theoretically impossible, all dictionaries are disguised encyclopedias. We must assume a more global knowledge is necessary if language is to be interpreted. How can global knowledge be represented so as to be discussed and theorized? All such representations are postulates and take the format of a multidimensional network (68).

[21] The representation Eco chooses for this network is a rhizomatic labyrinth. Rejecting first, the classical labyrinth of Crete, one in which you cannot help but reach the Minotaur at the center, and second, the Manneristic maze,3 a labyrinth which gives you choices, some of which lead to dead ends, in other words, one in which you can make mistakes, one in which "the Minotaur is the visitor's trial-and-error process,"4 Eco chooses a third type of labyrinth, the net, a labyrinth in which you cannot make mistakes since the point of such a net is to meander, to discover, to make connections.

The main feature of a net is that every point can be connected with every other point, and where the connections are not yet designed, they are, however, conceivable and designable. A net is an unlimited territory. (81)

[22] Further, Eco tells us that the best image for such a net is the rhizome suggested by Deleuze and Guattari (Rhizome). Such a net is like the rhizomes of the vegetable and fungal world. Some of its characteristics are: a. all points can and must be connected to all other points; b. it is anti-genealogical; c. it has neither an outside nor an inside because it makes another rhizome out of itself; d. it is susceptible to continual modification; e. one cannot provide a global description of the rhizome, not just because it is complicated but because it changes over time; f. there is the possibility of contradictory inferences because every node can be connected with every other node; g. it cannot be described globally; rather, it must be described as "a potential sum of local descriptions;" h. since it has no outside, it can only be viewed from the inside. A labyrinth of this kind is necessarily myopic since no one can have the global vision of all its possibilities, only the local vision of the closest ones. Because of this, every local description of the net is an hypothesis, "in a rhizome blindness is the only way of seeing, and thinking means to grope one's way (82)." Thinking means feeling our way along a local path which can change and change again at any moment. Instead of a static tree which disallows growth, we find ourselves in a universe of knowledge subject to continual revision and expansion. The tree is impoverished knowledge; the rhizome infinite possibility. Which matches our imaginings of literacy best?

[23] Finally, Eco tells us that "the universe of semiosis, that is, the universe of human culture, must be conceived" to be structured like the rhizome’s labyrinth. Every attempt to codify local knowledges as "unique and 'global' -- ignoring their partiality -- produces an ideological bias (83-84)." Such local knowledges have the potential to "be contradicted by alternative and equally 'local' cultural organizations (84)." And each of these claims it represents Truth. In other words, paradox becomes a familiar part of such a net. When local knowledges meet one another, paradox pops up. This can drive an individual into neurosis or start a war. But an acceptance of paradox as a condition of life can lead not to an interruption of ideology within its system, but a kind of truce between local knowledges, an agreement to disagree.

[24] Such a net represents an alternate to the dictionary and encyclopedia as a method for imagining literacy. It is the net to which this study turns later. Rhizomes are a powerful representation for an alternate conception of literacy, but for the moment, we are caught in a Porphyrian tree.