Chapter 25Lexical/Contrastive Accent and Minimal Pairs

25.1 Predictable accent and lexical exceptions

In Saru and other varieties with the usual accent pattern, a closed first syllable normally takes the accent; after an open first syllable, the second normally takes it. A substantial set of words departs from the second rule by having initial accent Nakagawa (2024: 50–51); Satō (2008: 15). These exceptions make accent part of the information needed to identify a word.

Nakagawa's Saru examples include húci “grandmother,” húre “be red,” mína “laugh,” núman “yesterday,” and tópen “be sweet.” The list is illustrative rather than exhaustive Nakagawa (2024: 51). The default rule alone would place the accent incorrectly in each case.

Satō also records rarer third-syllable accents in his Chitose description, including aokáy, taanúka, and toanúka Satō (2008: 15). These are lexical exceptions outside the Yakumo placement pattern described in Chapter 23 (The Pitch-Accent Placement Rule and the Accented/Accentless Dialect Split).

25.2 Words distinguished by accent

The clearest evidence comes from minimal pairs: the consonants and vowels are the same, and the accent position distinguishes the meanings. Nakagawa gives the following Saru pairs Nakagawa (2024: 48, 51). The English glosses below are supplied here.

The second member of each pair follows the usual rule for an open initial syllable. The first has exceptional initial accent. An unaccented spelling such as nisap consequently leaves information to be recovered from the context or the speaker's pronunciation.

Shiraishi gives another pair, citing Kindaichi and Chiri:

These pairs establish a lexical distinction in the described systems. They do not imply that every word has an accentual counterpart, or that the same pairs contrast in a variety without lexical accent.

25.3 How much work the contrast does

Shiraishi describes a small number of accent-only minimal pairs, while also documenting many words that follow the ordinary placement rule Shiraishi (2022: §5.1). The number of minimal pairs and the number of lexically exceptional accents are different quantities. An exceptional accent need not have a segmentally identical word with another accent.

The published pairs show that accent can distinguish words. They do not provide a measurement of how often listeners rely on that distinction in conversation. Such a measurement would also have to account for syntax and context: “suddenly” and “shin,” for example, usually occur in different grammatical environments.

Regional scope is essential. Nakagawa and Fukazawa distinguish accentual varieties from varieties in which pitch does not have this lexical function Nakagawa & Fukazawa (2022: §3.1). In Shizunai, Nakagawa reports variation between initial and second-syllable prominence in tokens meaning “suddenly”; this is part of his argument for an unaccented analysis, which differs from Refsing's description Nakagawa (2024: 50); Refsing (1986: §10.6) ‹contested›.

Accent marks are therefore most informative when their source and variety are known. A dictionary's notation can represent a particular speaker or regional pattern; it should not silently be imposed on every recording of the same word.

25.4 Complex words and related forms

The accent of a simple word does not always determine the accent of a larger expression. Shiraishi describes both preservation of first-element accent and renewed application of the usual placement rule in complex forms Shiraishi (2022: §5.2). The contrast between lexical accent and predictable placement must therefore be examined at the level of the complete form being pronounced.

A useful comparison keeps the base, the derived or repeated form, and their accents together. It also distinguishes an attested whole-word accent from a prediction based on the spelling. The relevant patterns are described in Chapter 26 (Accent in Compounds and under Affixation and Cliticization) and Chapter 32 (Reduplication and Phonological Alternations).

Sakhalin comparison. Some lexical exceptions have corresponding long vowels in Sakhalin: Hokkaido mína and Sakhalin miina “laugh,” for example Nakagawa (2024: 51). This comparative pattern helps identify a historical question; it does not turn the modern Hokkaido accent into a vowel-length distinction.