Chapter 101The Nominalizers -p/-pe and -i/-hi (Participant, Event, Place, Fact)

101.1 Form and place among the formal nouns

The nominalizer p ~ pe turns a finite clause into a noun phrase. It is a member of the closed class of formal nouns — light, weakly autonomous nouns placed after another element to build an NP — and it supplies a generic nominal head meaning, in its concrete uses, ‘the one who’ or ‘the thing that’. Hokkaido Ainu has no participial or otherwise non-finite verb form for this work: the verb inside the nominalized phrase keeps its ordinary person marking, polarity, number, and clause-internal morphology, and subordination is signalled by the following formal noun rather than by any change in the verb Bugaeva (2016: 100, 226–228); Nakagawa (2024: 503, 512–513); Satō (2008: 174–175).

The allomorphy is phonologically conditioned: the short form -p follows a stem-final vowel, and -pe follows a stem-final consonant. The same split holds whether the host is a single stative stem (poro-p ‘the big one’, pon-pe ‘the small one’) or a full clause (a=e p ‘what one eats’, okay pe ‘what one carries’) Satō (2008: 175); Refsing (1986: 260); Bugaeva (2022: ch. 14). Satō records a sandhi variant pa at the junction of p and a following conjunctive wa Satō (2008: 175).

The form spans lexical derivation and syntax without a break. Shibatani describes -p(e) as deriving a noun for a person or thing characterized by the verb, as in pirka-p ‘good thing’ and ray-pe ‘the dead’, and as a productive phrasal and clausal nominalizer besides Shibatani (1990: 33–34). In Bugaeva's areal account of nominalization, Hokkaido expresses clausal nominalization through a nominalizing word placed after the clause, the type [(np…) V p/pe], a stage from which lexical items such as aep ‘food’ have crystallized; the Sakhalin varieties extended the same system to possessed verbal forms Bugaeva (2016: 243, 385–397). Among the formal nouns, p ~ pe keeps a ‘thing/person’ profile beside the maximally abstract hi and the sensory-evidential heads ruwe, siri, hawe, and humi, whose inventory and possessed morphology are set out in Chapter 35 (The Formal / Defective Noun Set (ruwe, hawe, siri, humi, hi, pe, kur)) and whose clause-final evidential use is treated in Chapter 120 (The Nominalization-plus-Copula Evidential Schema).

101.2 Participant nominalization: the one who, the thing that

In its participant reading, p ~ pe is a generic head whose referent is fixed by the argument structure of the clause before it. The clause contains a gap, and the role of that gap — agent, patient, or an oblique participant — is recovered from the verb's valence and from any postposition left in place Nakagawa (2024: 512); Bugaeva (2015: 47). This is the headless edge of the relative-clause system: where an ordinary relative clause precedes a lexical head noun, p ~ pe supplies the head itself. The gap strategy, the accessibility hierarchy, and the postposition-retention facts are developed in Chapter 103 (Relative Clauses: The Gap Strategy and the Accessibility Hierarchy), and the place of the headless type within the wider typology of noun-modifying clauses in Chapter 104 (Headless, Internally-Headed, and Noun-Modifying Clause Analyses).

Lexicalized items preserve the syntax at word scale. aep ‘food’ is a=e-p, ‘thing one eats’, with the fourth-person transitive prefix and the nominalizer Bugaeva (2016: 245–251). Nakagawa's contrast between the synthesized a=e-suke-p ‘kitchen utensil’ and the compound sukemakiri ‘kitchen knife’ shows the diagnostic: the relative-clause nominalization keeps the applicative e- that licenses the instrument it names, whereas a compound leaves the relation between its two nouns grammatically unspecified Nakagawa (2024: 506–507). The lexicalized inventory is treated in Chapter 102 (Lexical-Head and Formal-Noun Nominalization (kur, uske, ruwe/hawe/siri/humi)).

An oblique participant is nominalized the same way, with the governing postposition retained inside the clause so that the relation between head and predicate stays overt. With an instrument the postposition ani ‘with’ remains, and p names the instrument Nakagawa (2024: 504); Bugaeva (2015: 47–49); (Keenan & Comrie 1977).

Ijäs teaches the same rule for relative clauses generally: the modifying verb phrase precedes the head, there is no relative pronoun, and an oblique head keeps the postposition fronted within the clause Ijäs (2023). Her examples are constructed teaching material and align with Nakagawa, Satō, and Bugaeva.

101.3 Event and fact nominalization

When the clause before it is already saturated — every argument the verbs require is present — p ~ pe nominalizes the event or fact as a whole and serves as a complementizer rather than a participant head. Nakagawa treats some formal nouns, p ~ pe among them, as filling this complementizing role; Satō draws the line explicitly between a nominalizer (名詞化辞) that turns a whole content clause into a noun phrase and a formal noun that heads a relative clause, the first allowing no reconstruction of an underlying sentence Nakagawa (2024: 503–509, 512–513); Satō (2008: 174–180).

In this use p ~ pe stands beside the other complementizing formal nouns — abstract hi and the sensory heads — which differ in the construal they impose. The complement-taking predicates and the control and raising diagnostics are treated in Chapter 106 (Nominalized Complements and Control/Raising); the future-oriented sequence kuni p in Chapter 107 (The kuni Complementizer and Purpose Complements); the quotative complementizer sekor and the indirect-speech markers in Chapter 105 (Quotative Complementation and Reported Speech (sekor)); and the clause-final evidential reflexes of the sensory heads in the inferential ruwe ne.

101.4 Arity and the boundary between the two uses

Nakagawa's valence arithmetic keeps the participant and complementizer uses from collapsing into one reading. A formal noun has arity −1, on a par with hi and uske; a clause that still requires a participant has positive arity, and the formal noun fills the open slot, giving a relative-clause NP; a clause of arity 0 is complete, and the formal noun nominalizes the whole of it, giving an event or fact nominal Nakagawa (2024: 261–262, 504, 507). The surface string [clause] p/pe is identical in both; the grammatical relation differs.

101.5 Analyses and open questions

The descriptions agree on the substance — p ~ pe builds lexical, phrasal, and clausal nominals, heads a generic participant in a relative clause, and nominalizes a saturated clause as an event or fact — and differ on the labels. Shibatani and Refsing use ‘nominalizer’ broadly; Satō separates the whole-clause nominalizer from the relative-clause-heading formal noun; Nakagawa makes the formal noun a word class and the complementizer a syntactic role available to some of its members; Bugaeva treats these light bound nouns as the generic heads of gapped relative clauses Shibatani (1990: 33–34); Refsing (1986: 259–264); Satō (2008: 174–175); Nakagawa (2024: 508); Bugaeva (2015: 47) ‹contested›.

The constituency of the nominalized clause plus a following copula — the frame [clause] p/pe ne — is the deeper disagreement, and it is the same one that runs through the evidential system. One line analyses the structure as biclausal, a formal or possessed noun heading the preceding clause and closed by the copula; another holds that the erstwhile noun is grammaticalizing toward a monoclausal complex predicate. No source read here offers a constituency proof that decides between them Tamura (2000: 227); Satō (2008: 175); Bugaeva (2016: 306–315); Dal Corso (2018: 24–25) ‹contested›.

Ijäs's observations on relative clauses fit this system, with two cautions. Her report that the subject of a ditransitive resists being the head matches Bugaeva's finding that no example of an A-of-three-place-verb head is attested; the construction is an attested gap rather than a settled ungrammaticality ‹corpus-confirmed›. Her difficulty in finding ablative wa(no) ‘from which’ or mutative ne ‘as which’ as relative-clause heads is shared by the sources read here, so the forms she suggests for them are left aside rather than presented as established ‹corpus-suggested› Ijäs (2023); Bugaeva (2015: 48).

The same nominalizer feeds clause linkage at the far end of the cline: a bleached p ~ pe works as an adversative conjunction ‘though, and yet’, no longer an argument of the following clause, and the Horobetsu nominalizers grammaticalize a step further into adverbial-clause markers Nakagawa (2024: 515–516); Takahashi (2018: 107–112).

Token frequency confirms that pe is common but cannot measure this construction directly. The corpus word-frequency table returns 13,249 tokens of pe (rank 15), a figure that folds together the formal noun, the numeral counter -pe, and lexical pe ‘water’, while the short allomorph p is obscured by segmentation. A per-function count would require construction-level annotation ‹corpus-suggested›.