Chapter 52Relational Nouns, Spatial Forms, and Quasi-Incorporation

52.1 Relational nouns and grammatical place

Locative or relational nouns describe a position with respect to a reference point: a top, side, interior, or direction. They combine with place particles and have forms whose distribution differs from ordinary possessive nouns Nakagawa (2024: 239–241); Nakagawa (2022: §5). Their reference point can be an object, a person, or a location; it need not be an owner.

Nakagawa calls these expressions grammatical places because they can take place particles directly. An ordinary noun such as nupuri “mountain” normally needs a relational noun in the corresponding place expression, whereas kim “inland/mountains” already has place status Nakagawa (2024: 246–247). The relevant contrast is grammatical, even when both expressions refer to familiar locations.

52.2 Two classes and their independence

Nakagawa distinguishes two classes. In the first, a basic form needs an immediately preceding reference-point phrase or a person prefix. Its long form can instead recover the reference point from context. His comparison uses basic ka after “horse” and long kasi referring back to a dish mentioned earlier Nakagawa (2024: 239–241).

The second class can stand independently even in its basic form. It includes soy “outside,” aw “inside,” kim “inland,” pis “toward the shore,” rep “offshore,” and ya “shore/land,” among others Nakagawa (2024: 245–246, table 53).

Some authors classify members of this second group as ordinary nouns. Nakagawa includes them among locative nouns because of their long-form morphology and their place syntax Nakagawa (2024: 246) ‹contested›.

52.3 Person marking identifies the reference point

Locative nouns use the O series of person markers. Satō's Chitose paradigm gives the following forms; his inclusive label corresponds to the fourth-person series used here Satō (2008: 160); Nakagawa (2022: §5.2, table 5).

Both basic and long forms can occur with person marking, but the available combinations are lexically restricted. Satō gives Chitose en-sam and en-or, rejecting *en-sama and *en-oro. With “front,” he permits both en-etok ta and en-etoko ta Satō (2008: 165). The general marker series should therefore be kept separate from each noun's permitted word forms.

Person-marked examples from Nakagawa's second class are fewer. He records i=soyke, i=piske, and i=rorkehe, while treating wider availability cautiously Nakagawa (2024: 246, examples 302–304).

52.4 Basic and long forms

Nakagawa uses “basic” and “long” because the forms differ from the concept and affiliative forms of ordinary nouns in both morphology and syntax. Satō uses affiliative terminology in his treatment Nakagawa (2024: 239–240); Satō (2008: 160–165). Selected documented forms illustrate the range Nakagawa (2024: 241–245, 247–249):

The endings include vowel extensions, -si, -ke, and further long-form extensions. -Ke may attach directly to a basic form or to an already extended form. The patterns are lexical: the Handbook account, for example, excludes the -ke series for etok Nakagawa (2022: §§5.1, 5.3.1). The suffix also occurs beyond locative nouns, as in hike and aktonoke Nakagawa (2022: §5.1).

52.5 Spatial contrasts within the class

Several English spatial labels conceal distinctions made by the Ainu forms. Nakagawa distinguishes ka, above in contact, from enka, above with separation, and kurka, an extended upper surface Nakagawa (2024: 242–243).

For front and back, etok and oka orient a moving entity, a line, or time; kotca and osmak describe the front and back of something stationary. For interiors, onnay denotes a hollow space and tum a homogeneous interior. Oske covers both in the described distribution, with regional differences in the use of these words Nakagawa (2024: 243–245).

The basic forms therefore contribute spatial structure before a following particle specifies location, source, or direction. See Chapter 53 (Place Particles and Motion Expressions).

52.6 Or ta, otta, and dialect-specific independence

The sequence or ta can undergo the sound alternation represented as otta. Nakagawa also reports an independent otta in Shizunai and Tokachi, where the reference point occurs earlier in the discourse. In the corresponding Saru and Chitose environment he requires long-form oro ta Nakagawa (2024: 240, 297).

He analyzes the independent combinations as lexicalized postposed adverbs, rather than treating the behavior as a property of every basic locative noun Nakagawa (2024: 240) ‹contested›. Pronunciation, spelling, and syntactic independence are separate questions in this comparison; a written space alone cannot settle them.

52.7 Satō's quasi-incorporation analysis

The O-series marking on locative nouns raises a grammatical question: why does a reference point receive the same person form as a transitive object? Satō connects it to the noun–verb pattern he calls quasi-incorporation, distinguishing a morphological word from a lexical word Satō (2021: 89–94). The latter is a conventional unit of meaning that can retain the morphological structure of a phrase.

The verbal pattern is also described by Tamura as 連他動詞, a two-word verbal phrase functioning as a unit. Her account places O-series marking on the noun and A-series marking on the verb Tamura (1997: 15, §C.4). Satō's Chitose example makes that division visible:

Ordinary “my mouth” is ku-paro, with possessive A-series marking and the affiliative noun. In en-par a-oyki, basic par combines with O-series en-. Satō analyzes the possessor as promoted to the object of the whole lexical unit “support.” Morphologically, the expression remains two words. With derivational reflexive yay-, his yay-par-oyki is instead a single morphological word Satō (2021: 90–94).

Satō proposes a parallel analysis of or ta. In en-or ta “at my place,” the reference point receives O-series marking because it is treated as an object of the lexical unit; the locative noun retains basic or. The noun and case particle nevertheless form a morphological phrase. This extension of quasi-incorporation is a hypothesis about their combined structure Satō (2021: 91–94) ‹speculative›.

For cise or ta “at the house,” Satō further proposes incorporation of cise into the locative unit. His preferred analysis is represented as cise-or ta, but he also considers other segmentations. Specific phrases such as “this hole” before or complicate an ordinary incorporation analysis Satō (2021: 94–95, note 22) ‹speculative›.

52.8 Or, oro, and the limits of the proposal

An ordinary animate reference point favors affiliative oro in Satō's Chitose material: unarpe oro ta means “at aunt's place.” He links this to the difficulty of incorporating a specific animate participant. The full phrase can still count as a lexical unit in his analysis Satō (2021: 93–95) ‹speculative›.

The observed distribution has exceptions. Basic or occurs after words for gods, people, and doctors; Satō interprets these examples as generalized or occupational reference. His bear-agent examples contain both or wa and oro wa Satō (2021: 85–86, examples 11–15). Conversely, the inanimate phrase poronno an pe oro wa, “from among many things,” contains oro Satō (2021: 88, example 20; 92, note 20).

Affiliative morphology also occurs in verbal quasi-incorporation. Satō's sentence meaning “I intend to support my uncle” has an external possessor phrase and affiliative paro. He treats its conventional meaning as grounds for lexical unity despite the fully phrasal morphology Satō (2021: 90–91, example 23; 94). Thus, quasi-incorporation in this account includes more than the basic-noun pattern illustrated by en-par.

Satō acknowledges an unresolved difference between the two constructions: the inanimate-noun-plus-basic-locative pattern of cise or ta lacks a matching established pattern in his verbal quasi-incorporation examples Satō (2021: 95–96). He also sets aside orke because of limited evidence Satō (2021: 81). The proposal connects several otherwise separate facts about person marking, noun forms, and lexical meaning; these remaining differences constrain its scope.