Spoken artifacts

This morning I was finishing up reading Rorty’s electrifyingly provocative essay on pan-relationalism while sporadically talking with Susan about how she’s feeling stuck in her current project.

As always happens when I read Rorty, I’ve been thinking about words as one kind of artifact that humans devise and use, but which, to us humans (especially to Rorty!) seems somehow more fundamental and unique to who we are than all those other non-linguistic artifacts we make and use.

I suggested to Susan that she could try prototyping her idea. Then I realized people never understand what I mean when I say prototype. They imagine something much more finished, where in design a prototype is the fastest and cheapest way to get an idea out of one’s head and into form where it can be interacted with, so designers can learn from the externalized interaction and further develop the idea. To give her an idea of exactly how crude, I showed her the classic IDEO example.

Suddenly, I understood something new.

Language is a prototyping tool.

When we propose, suggest, speculate, plan we are prototyping possibilities using spoken artifact that we and others imagine, consider, try out, or try on. They, in turn, can descriptively manipulate our prototype, and present it back to us in modified form. Conversation of this kind is collaborative language prototyping.

Of course, language is not always the best prototyping material. As an idea progresses toward concrete actualization, it becomes more and more risky relying on what we learn from our spoken artifacts (unless, of course, the ultimate artifact is itself linguistic). The lessons gleaned from the verbal artifact may not apply as expected to the material artifact that is our “final product”. But this is true of any prototype. Proper use of a prototype always relies on interpretation, analogy and imagination. Failure to see the intended analogy between the prototype and the intended final artifact is similar to missing the intended meaning of a sentence.

This is why, after talking out an idea, then drawing it or acting it out, it prototypes are developed iteratively, in increasing levels of fidelity that successively approach the final product — but the whole time designers rely on language to prototype potential variants, before committing more resources. “Talk is cheap.” And this is what makes it such a great prototyping material!


For me, talking, imagining, drawing, crafting are all products of nonverbal intention — intuition.

Yes, these things are tough to talk about clearly, and if talking as clearly as possible is your goal, maybe you’ll want to deemphasize the role of tacit intention in your descriptions of how humans think, perceive, value and behave.

But if you are a designer, and your goal is to produce artifacts (including verbal artifacts!) that people find desirable, useful and usable and you discover that usability in particular requires a direct coupling between intention and action without intervention of language, you’ll find conflations of intention and language counterproductive, and clarity gained in confusing the two to be more of a vice than a virtue.


I seriously want to make a linocut print of the IDEO prototype.


I wrote something last week in an email that I want to glue to this line of thought.

…I have a redescription of panrelationalism to try out on you — one that does more justice to tacit knowledge. I’m happy to call “tacit knowledge” an ability and not a form of knowledge, as long as we distinguish ability from instinct or reflex by specifying that an ability is an intentional act situated within a pragmatic network of consequential (often unstated) beliefs. 

I see language use as a special case of ability, the ability to use a certain set of words — but that the ability to wordlessly do something with an intent — “intentional attitude” should take the place of the “sentential attitude.” A sentential attitude is one variety of intentional attitude, an intention to speak. 

We do not have to form a sentence to have an intention, but every sentence is directed by an intention, which is why we discover the ending of sentences as we utter the last word of it. We are using a tacit ability as we interact with an already-said sentence and available vocabulary for extending it, guided by a tacit intuition of what we want to say, in the very act of speaking. 

If we replace every linguistic term with an intentional one, and understand that language is a special case of intention, I’m a panrelationalist. I’ll even extend an olive branch and admit that our linguistic genius can, in principle, speak any intention. 

But the tradeoffs involved in allowing words to metonymically stand in for intentions is unacceptable for this designer, since good design is so closely bound up with wordless intentional actions. The metonymy erases a crucial distinction designers need to make. But also, there’s a fine line between metonymy and conflation, and I think philosophy is best served when we split hairs when we can show that there is a difference there that makes a difference. I think the “good design” difference is only the tip of the iceberg of reconceiving and potentially redescribing how minds use language and other tools.

1 thought on “Spoken artifacts

  1. As usual, your brilliant metaphors bear incredibly fruitful epiphanies for me. Your metaphor of language as a prototyping medium caused me to go back and read Reddy’s “The Toolmaker’s Paradigm” again. As is my habit, I followed up my rereading with a google search to see if the essay had borne any new fruit. I found an orchard in the form of a book: “The Instruction of Imagination” by Daniel Dor (2015). Inspired by Reddy’s metaphor, Dor has created a theory of language and its evolution that is incredibly useful, usable, and desirable! I’ll send you details by email, but for your general readership I’ll briefly redescribe Dor’s theory as follows.

    Language is a prototyping tool that enables a speaker who wants to convey a conceptual experience from their mind to create a message that acts as a set of instructions to the listener for prototyping an approximately similar conceptual experience in their mind.

    Here’s how Dor puts it:
    «Language resides in the social domain. It allows us to systematically channel complex sets of instructions for imagination from mind to mind, across the experiential gaps that separate us, through the socially constructed worldview of the mutually identified symbolic landscape. Our words function as discrete instructors, superimposing sets of semantic categorizations on our private, variable and analogue, experiential worlds.»

    Dor reestablished personal experience as the private, ineffable source of the meaning of our utterances. There is no such thing as a private language, but language exists in order to narrow the “experiential gap” between interlocutors’ private experiential imaginations:
    «This is precisely why the entire process of linguistic communication is so fragile. Understanding depends on our ability to go beyond the conventions of language and re-enter the private domain of experience, but the very fact of the experiential gap implies that whatever we do, we always re-enter our own experiential world, never that of the speaker. Every understanding is also, at the very same time, a misunderstanding.»

Leave a Reply