Reading Heidegger’s What Is Philosophy? set me going. Heidegger answers the question not by definition but by retrieval: he takes philosophia back into its Greek setting, as though the older word might disclose what philosophy is. The lecture is the occasion rather than the target. It offers a clean specimen of a manoeuvre that runs far wider. A word’s origin can show where a practice has been, but it can’t settle what the practice is. At least this is what I was thinking as he was making his case.
Four distinctions keep that confusion from forming, each marking a method mistaken for the disclosure it can’t perform.
Etymology is not ontology. A word’s origin gives a genealogy, not an essence.
Semantics isn’t phenomenology. Describing the uses of ‘philosophy’ isn’t describing the activity, pressure, or encounter that those uses later name.
A definition isn’t an encounter. To inventory a term from the outside isn’t to be drawn into what it names; philosophy becomes a drawer to be catalogued rather than a movement to be entered. Consider the affordances of McGilchrist’s left-cerebral hemisphere metaphors.
Semiotics isn’t adjudication. Whilst it can map how a sign orients attention, circulates, and acquires force within a field, it can’t convert that orientation into a standard against which correct use might finally be judged. The sign is a flag in a nebula, rather than the ground beneath it, as it were.
What these distinctions share is a single diagnosis. For a wide class of terms – ‘philosophy’ among them, alongside truth, justice, freedom, consciousness, Being, and the rest of the overpromoted vocabulary – there is no referential anchor stable enough, and no court neutral enough, to adjudicate correct application without appeal to the contingent history by which the term became usable. In my language insufficiency hypothesis, I mark the position such a term would occupy, were its use settled frame-independently, as an Invariant: not a discovered class here, but a limit-position against which instability becomes visible. Wherever correct use depends on a stabilisation history that could have gone otherwise, rather than on a frame-independent anchor, the term is, rather inelegantly, non-Invariant. I used to say weasel word, but that doesn’t capture enough of the class to be useful. Inquiry doesn’t terminate against such terms. It maps the conditions of their temporary stability.
The demand that inquiry ‘gets to the bottom’ of such terms, therefore, fails by misconstruing the kind of object under examination. There is no bottom to reach if ‘bottom’ means an essence beneath usage, history, practice, and contestation. (Don’t force me into defending Dasein.) What can be recovered are sedimented uses, institutional pressures, inherited distinctions, pragmatic settlements, and the conditions under which a term becomes authoritative enough to govern discourse while remaining unstable enough to resist closure. An inquiry doesn’t discover a hidden foundation. Rather, it discloses why the demand for one was misplaced.
Paradox Lost
To be honest, the paradox isn’t lost on me. Wittgenstein reached it first: the critique is conducted in the very medium it indicts, and there is no metalanguage exempt from the diagnosis. ‘Frame-independent’, ‘neutral’, ‘stabilisation history’, and ‘non-Invariant’ are themselves terms without a neutral court. But this is the ladder, not the contradiction. The recursion doesn’t embarrass the position; it demonstrates it. A diagnosis that exempts its own instruments would be making the very claim to bedrock it denies everyone else: Heidegger’s error wrapped in foil. The point isn’t that nothing is there, but that nothing adjudicating is neutral. This isn’t anti-ontology. It’s the refusal to support one prematurely.
Enfin
This position isn’t offered as an immaculate invention, ex machina, sprouting whole from the head of Zeus. It stands in a lineage of suspicion toward linguistic authority. Saussure makes visible the differential and conventional character of the sign; Wittgenstein displaces essence into use and family resemblance; Heidegger exposes the way inherited metaphysical vocabularies conceal the question they claim to ask; Derrida radicalises the instability of presence and the deferral of meaning; Nietzsche treats concepts as hardened metaphors whose genealogy has been forgotten; Gadamer reminds us that understanding is historically situated before it is self-possessed; Foucault shows how discursive formations are stabilised through institutional power; Peirce, from another semiotic route, treats signs not as inert labels but as interpretive processes. Gallie’s account of essentially contested concepts gives one formal ancestor for the non-Invariant region, though the present claim is broader: the issue is not merely that some concepts are disputed, but that their apparent adjudicability often depends on stabilisation histories mistaken for foundations.
In humility, the present nomenclature isn’t an attempt to escape these thinkers, still less to supersede them by terminological vanity. It’s an attempt to compress several inherited insights into a diagnostic vocabulary suited to a specific problem: the tendency to mistake semantic, etymological, semiotic, or definitional stabilisation for ontological disclosure.