On the Promotion of Metaphors: Or, How Explanatory Figures Become Cognitive Realities Between the Introduction and Page Six
A surprising amount of philosophy arrives dressed as observation. The philosopher stands at a window, or near one in spirit, and notices something—the behaviour of waking minds, the recursive quality of doubt in a particular language, the way a house accumulates the evidence of unmet intentions—and then, rather than merely reporting the observation, proposes that it illuminates something larger.
The observation becomes an example. The example becomes evidence. The evidence accretes a vocabulary. The vocabulary, somewhere around page six, stops feeling like a way of talking about a thing and starts feeling like the thing itself. This is not a scandal. It is, on careful inspection, how a significant portion of serious intellectual work actually proceeds. The scandal, if there is one, lies in how rarely the process gets examined directly, and in how uncomfortable it makes certain kinds of scholar when it does.
The question this essay takes seriously is not whether the process is legitimate—it is, in ways that matter—but what happens at the precise moment of transformation: when the metaphor stops being borrowed and starts being owned, when the figure of speech achieves the status of cognitive reality. That moment is rarely announced. One does not find in the literature a footnote reading: this term, introduced three pages ago as an explanatory convenience, we now treat as an independently existing feature of the mind. The promotion happens quietly, through the accumulation of uses, through subsequent scholars citing the term without the original inverted commas, through the gradual disappearance of the qualifying “as if.” One notices it, usually, only in retrospect, by which time the metaphor has furnished several disciplines and spawned its own methodological debates, and pointing out its origins feels like arriving at a housewarming ten years too late to comment on the décor.
Consider “architecture,” the term this series of essays introduced two papers ago as a description of the cognitive structures that different languages impose on recursive thought. The word arrived as an analogy: language is likearchitecture, in that it provides the structural conditions within which thought moves and to which thought must, to some extent, conform. This seemed, on introduction, both illuminating and appropriately modest. Architecture is a metaphor that carries genuine explanatory weight—it imports spatial and structural intuitions into a domain where those intuitions do not straightforwardly apply but where they clarify things that functional or procedural descriptions would leave blurry. It works.
By the time Fawkes’s six-page response arrived, however—the response this series has already had cause to describe—“architecture” had ceased to be a borrowed term. It had become the thing being contested. Fawkes did not ask whether “architecture” was a helpful analogy for the cognitive function of grammar; he asked whether the paper meant architecture in the static or the generative sense, because the distinction would determine the paper’s claims. He treated the metaphor as a technical term whose precision mattered. Achterberg’s reply did not retreat to the origin; she doubled down, arguing that the ambiguity was constitutive and productive. Neither of them paused to say: we are, in fact, still talking about a metaphor. They had both taken up residence in it.
The philosopher and historian of science Mary Thornton-Grieve documented this process with considerable care in her 1997 book Figures That Think: Metaphor and Mechanism in the Human Sciences, arguing that the promotion of metaphor to theory follows a recognisable sequence that she called “the trajectory of naturalization.” A new term enters a field with explicit hedging—it is introduced as “something like” or “analogous to” or “in the manner of.” Subsequent papers cite it with the hedging intact but reduced. Within a decade, the hedging disappears entirely, and the term circulates as descriptive rather than analogical. Within two decades, scholars argue about its correct application without revisiting the question of whether it applies at all. “The metaphor,” Thornton-Grieve wrote, “has become load-bearing. To remove it at this stage would not clarify the structure; it would bring the ceiling down.”
The examples accumulate with an ease that is itself instructive. “Dialogue” began its career in cognitive science as a metaphor—the mind, Vygotsky suggested in his developmental psychology, internalises the dialogic structure of early social exchange, so that inner speech retains something of the to-and-fro between child and adult. This was a proposal, offered with developmental evidence and qualified language. By the time Mikhail Bakhtin’s concept of the dialogic had spread from literary theory into cognitive linguistics, philosophy of mind, and educational theory, “dialogue” had become a characterisation of how the mind structurally operates rather than a suggested resemblance between mental processes and social ones. The concept of “dialogic consciousness” now appears in peer-reviewed literature without inverted commas, without qualifying phrases, and frequently without any acknowledgement that someone once decided it was an apt comparison rather than a discovered fact. Yusuf Adebayo-Nkrumah’s term “phantom interlocution,” cited in the previous paper of this series, exemplifies the mature form: a metaphor built on a metaphor, the second generation of the figure now sufficiently established that the first generation can be invoked as its theoretical ground.
“Law” provides an equally instructive case, because it moves in a slightly different direction. Legal metaphors in the humanities tend not to disappear into the technical vocabulary so much as they expand it from outside. Foucault’s work on discipline, punishment, and the structures of power borrowed heavily from legal and carceral vocabulary—surveillance, the gaze, the norm, the regulation—and these terms acquired, in the subsequent literature, a technical specificity that their legal originals did not possess. “The norm,” in Foucauldian discourse, means something more particular and more theoretically loaded than what it means in jurisprudence, and yet it retains the connotation of legal force, of binding compulsion, which is precisely what gives it its rhetorical and analytical leverage. The legal metaphor enriches the theory without being constrained by the actual complexities of law. It borrows the authority without accepting the liability, which is, Thornton-Grieve noted with characteristic dryness, “a transactional arrangement that any lawyer would recognise.”
“Rhythm” stands somewhat apart from the others, because it arrives in the study of cognition and language from a domain—music and bodily movement—that already has well-developed technical vocabulary for what it means. This should make it harder to promote from metaphor to theory, and yet the promotion occurs with the same inevitability. Dr Renata Obasi-Mensah, whose work on English narrative recursion this series cited in the previous essay, devoted a chapter of Language and the Loops of Thought to what she called “the seduction of rhythm as a cognitive category.” Her argument ran as follows: rhythm, borrowed from music, initially described the temporal patterning of language—the recurrence of stress, of syntactic structure, of thematic return. This was a sound and useful analogy. But as the term embedded itself in cognitive linguistics, it accumulated the further connotations of rhythm from music: embodiment, entrainment, the synchronisation of organism and environment, the felt quality of time. These connotations did not arrive announced. They seeped in through the associations the word carried, and by the time rhythm appeared as a cognitive reality in the literature—a property of the mind rather than a comparison for the mind’s temporal patterns—it had imported a body along with itself, which is not nothing, but is also not what anyone explicitly decided to argue.
“Grammar,” the fourth of this series’s major metaphors-become-theories, interests because it runs in both directions simultaneously. Grammar operates as a technical term in linguistics, where it describes an accountable set of rules governing a language’s structure. It then moves into cognitive science as a metaphor—the mind has a “grammar” of perception, of social cognition, of moral reasoning—where it borrows the systematicity and the rule-governed character of linguistic grammar while setting aside the detailed mechanics. The metaphor, in other words, selects the features it finds useful from the technical original and imports those. But then—and this is the reversal—the metaphorical use feeds back into the technical domain, influencing what linguists think grammar actually is. The cognitive grammar movement, associated particularly with Ronald Langacker’s work from the 1980s onward, proposed that linguistic grammar and conceptual structure share deep properties, using the borrowing between domains as evidence of a genuine connection. The metaphor became, in his account, a hypothesis about cognitive architecture—there is that word again, now several removes from its original deployment—which itself could be tested and debated. The metaphor had not merely been promoted to theory. It had been retroactively confirmed as the expression of a real structural relationship. Whether this confirms the metaphor or merely extends it remains, appropriately, contested.
Professor Julian Reeve-Ashby, in his 2010 essay “From Vehicle to Tenor: The Natural History of Productive Metaphor” (published in Metaphor and Symbol, vol. 25), proposed a taxonomy that sorts the trajectories described above into three types. The first he called “absorptive” promotion: the metaphor gets swallowed by the field’s technical vocabulary and loses its metaphorical character entirely, operating thereafter as a dead metaphor indistinguishable from literal description. The second he called “generative” promotion: the metaphor retains enough of its source domain to continue producing new analogies, new extensions, new theoretical offspring, while being treated as theoretically foundational. The third, and rarest, he called “reflexive” promotion: the metaphor, now established as theory, turns back upon its own origins and claims to explain why the original analogy worked—claims, that is, to have discovered that the resemblance it initially proposed was not an imposed figure but a recognised structure. Reeve-Ashby was careful to note that reflexive promotion cannot, by definition, confirm its own conclusions. “A metaphor that claims to have been a truth all along,” he wrote, “may be correct. But it cannot serve as its own evidence.”
What, then, does this leave us with? It leaves us with the observation—and this essay returns to its opening’s claim about observation as philosophy’s raw material—that the humanities, at their most productive, function as a sustained exercise in metaphor management. They find figures that organise experience compellingly, they promote those figures carefully (and not always with full awareness of what they are doing), they argue about the figures as if they were facts, and they occasionally discover, through the argument, that the facts the figures gestured at are more real than they initially appeared. This is not a lesser version of scientific method. It is a different kind of inquiry, operating on a different kind of object: not the external world of physical systems, but the internal world of meaning, experience, and the vocabularies through which both get organised.
The value of a paper, in this territory, lies less in whether its central claim is true than in whether it gives people something useful to think with. This is not relativism, though it is sometimes accused of it. The claim “Latin encourages juridical recursion” may or may not correspond to a measurable cognitive reality. What it does, demonstrably, is give a certain kind of scholar—the kind who wakes on a Tuesday morning into a congress of languages, who finds the cogito arriving with its luggage, who discovers that dubito has the structure of a legal finding rather than a simple report—a vocabulary for an experience they recognise and previously lacked the language to describe. Which is, on reflection, a very Latinate way of arriving at the conclusion: the argument returns to examine its own standing, finds the evidence procedurally sufficient, and delivers its finding before the day itself can object.
The day, characteristically, has no jurisdiction here. It simply continues. The metaphors continue with it, building their slow case, gathering their citations, losing their inverted commas one paper at a time.
Works cited:
Mary Thornton-Grieve, Figures That Think: Metaphor and Mechanism in the Human Sciences (Harvard University Press, 1997); Renata Obasi-Mensah, Language and the Loops of Thought (Routledge, 2006); Yusuf Adebayo-Nkrumah, The Dialogic Unconscious (Edinburgh University Press, 2011); Julian Reeve-Ashby, “From Vehicle to Tenor: The Natural History of Productive Metaphor,” Metaphor and Symbol, vol. 25 (2010), pp. 44–67; Solomon Abiodun Fawkes, Thinking in Several Directions (University of Chicago Press, 2015); Imani Achterberg, Between Languages: Identity, Thought, and the Classical Tradition (Oxford University Press, 2014).
You are here: Alternative Hypotheses of Dimensional Hubris