Several things came together for me when I woke up this morning. A central problem of considering photographs to be components of a “language” is that they are unique. A photograph can substitute for a thing, but it resists consideration as a representation of type of thing. The situation is analogous to proper names and generic names: a photograph, in conventional use, is generally treated as a proper name rather than a generic label. As a predicate, it is never necessary but rather posits a metonymic connection with the thing it identifies. Metonymy can only be broken by surrendering the photograph’s status as a “true” or “proper” representation of this thing—through its conversion into a metaphor. This conversion comes from the attribution of a metaphoric significance—an implicit cognitive comparison that stretches the boundary of the proper name.

Under this lens, photographs are nouns. However, the metaphoric attribution of “feelings” or “actions” to the lifeless images can transform them into verbs, or adjectives and adverbs, making propositions possible. Extracting “meaning” from images seems to be inextricably tied both to the process of metaphor and to the process of translation. The status of photographs as semic units is mired in a swamp of contentions about their ability to contain codes that can be processed and decoded: is there such a thing as photographic diction? Do images have a grammar, or at the very least a syntax? Are images nouns, phrases, clauses, or can they be propositions? Stepping up another level, can a photograph present an argument or proof?

Clearly, the translation of photographs to test this sort of assertion is a hermeneutic process which crosses the border from diction into thought. This is precisely the boundary that Aristotle chose not to cross in his discussion of metaphor in the Rhetoric and the Poetics. Metaphor is discussed as a figure of diction, not of thought. At the level of diction, metaphor is defined as a transgressive attribution—the application of an “improper” name to a thing. For photographs to work as a language, such an attribution is essential. At the phrase or clause level, such a figure could also be labeled as catachresis.

Catachresis is the trope of absurd comparison: God is a fisherman, or if you’re familiar with Concrete Blonde, God is a Bullet. However, catachresis can also be used to invent terms that did not exist before. A catachresis is often used to productively refer to a “thing” that exists outside of language—something that has no “proper” name. Thus, catachresis is essentially untranslatable given the common frameworks of interlingual and intralingual translation. However, following I.A. Richard’s tantalizing suggestion that metaphor is an intersemiotic translation (appropriating the term from Roman Jakobson), untranslatable is actually a misnomer. What seems to be at work instead, is a transgression of boundaries between languages— in the verbal examples I’ve given, it is a transgression between the boundary between words and thoughts.

To translate a photograph into words (interlingually) or a metaphor into paraphrase (intralingually) represents only a limited glimpse into the possibilities of translation. To transgress the boundary between diction and thought does not necessarily assume that both are forms of language—this would be an interlingual (traditional translation) problem. It is equally possible that the translation which occurs when we “read” a photograph or conceptualize a metaphor happens between semic systems that are not necessarily compatible in features or transposable in their parts, except as an incredibly weak analogy.

Is thought a language? Perhaps, but perhaps not.