A Grammatical Appendix

1 Preliminaries

Periodisation

Old Chinese and Middle Chinese, like Baroque, are terms that one could wish by now universally discredited. It may be recalled that Karlgren’s Ancient and Archaic Chinese were never intended for periodisation.

Use a date, date range or political period.

Example: Early Chou Chinese

2 Syntax

Gaps in terminology

See further On Diglossia and consistency of sets.

Emic

The consequences of a failure to distinguish between morphs and morphemes are of two kinds:

1. By using the term morpheme when they really mean a morph, one can like Huang–Shi miss a whole phenomenon like orthographic alternation.
2. By not grasping the difference between abstract units and their physical realisation, one can fall prey to irrelevant concepts such as erh tzu tz’u and shuang yin chieh tz’u which describe orthography and pronunciation respectively.

4 The clause: complements

This is a chapter omitted by Huang–Shi, to its own detriment.

Copula clauses and verbless clauses

In the declarative mood, an adjectival CC requires a degree or negative copula as predicate. Zero surface realisation of the predicate is not allowed. Cf. *選舉重要.

In the interrogative mood, zero surface realisation of the predicate is allowed.

6 Adjectives and adverbs

[顯然][是]__

Without testing this collocation with paradigmatic alternants of different word classes in the slot, Huang–Shi §12.2 jump to the conclusion that 顯然是 is an adverb. Compare, however:

8 The clause: adjuncts

越- PPs (classical)

15 Coordination

Asyndetic coordination

Dictionaries and Chang §4.4.4.3 give the impression that as a marker of asyndetic coordination -等 must be followed by a numeral classifier. This is not true; other clues, e.g. 都, can also serve to show that it is an exhaustive enumeration.

Example: 比如時間、地點、工具等短語都有語序變化問題 (2010)

19 Lexical word-formation

False tests

Expansion-based tests are not a reliable tool for distinguishing between syntactic constructions and compounds. Their operation ultimately depends on knowledge of semantic content and orthographic conventions. Under the delusion that syntactic rules are innate, however, faulty lexical analysis is often used to explain away conflicting evidence.

Examples: A matchbox is not a box for matches (CGEL §14.4), but one in which matches are sold, with a striking surface on one side; 羊肉 is not equivalent to 羊的肉 (Chang §5.2.2.3), but flesh of sheep used as food.

Two-character morphs

-主義 (1883) was borrowed from Japanese as a bound root. It constitutes a single morph which cannot be analysed into 主 and 義. Similarly, -製品 (1880).

Defining ‘word’

In Mandarin Chinese, a native word is a lexeme realised by

1. a free morph or
2. a fossilised compound of two or more morphs or
3. a compound of a bound root as head and a word as dependent or
4. a fossilised NP.

Words vs phrases

Huang–Shi §8.5.1 gives up trying to distinguish between compounds and NPs; Chang §5.2.1.2 meekly accepts Lü’s arbitrary rule. The following analysis implements the defition above on Chang’s examples:

Implementation for examples in Huang–Shi §3.4.2:

Orthographic alternation

A lexeme is often realised by multiple orthographic variants, which can be either stylistic variants or free variants. The alternation takes the form of allomorphy in one or more of the morphs.

Example: 通假 (s.xii) vs 通叚 (s.xix)

-等國

Huang–Shi §3.5.1.1:

英法等國 … ‘countries like England and France.’ … a case of conventionalized contextual ellipsis.

Both the gloss and the analysis are wrong. They illustrate some of the commonest pitfalls in synchronic analysis:

1. The elided part must be unambiguously specified.
2. No attention is paid to paradigmatic alternants like -等地, where a semantically equivalent full form is not in use. Cf. 英法等國 vs 英法等國家 with 台北上海等地 vs ^#台北上海等地區 vs ^#台北上海等地方 vs 台北上海等城市.
3. No attention is paid to more established alternants such as 各國 (1602) and 各地 (1822), from which -等國 and -等地 are created by analogy.

英法等國 should be glossed as ‘Britain and France’ (等 as marker of asyndetic coordination) or ‘Britain, France and other countries’ (等 as abbreviatory device).

20 Punctuation

By abandoning any pretence of being descriptive of real-world usage, the appendix on punctuation is probably the weakest chapter in Huang–Shi.

An adequate treatment of Mandarin punctuation must take account of the following facts:

1. Punctuation marks are not an integral part of written Mandarin. It is very common in written communications to use space as the only punctuation indicator.
2. The actual set of punctuation marks being used varies considerably depending on the particular Mandarin-speaking region and personal preferences as influenced by contact with other languages, especially Japanese and English.
3. Unicode glyphs and descriptions must be used if punctuation marks are to specified unambigously.