© SHESL/EDP Sciences

The first Westerners to learn and describe Cantonese language were mostly Protestant missionaries who arrived in China in the 19^th century. In order to meet their own communicative needs and those of the merchants they often accompanied, they began to present and translate Cantonese language in grammars, dictionaries, primers and phrasebooks, also devising Romanization systems to transcribe its sounds. The first attempt at describing Cantonese grammar was made by Robert Morrison (1782-1834) − the founder of the Protestant missions in China − in a section of his Grammar of the Chinese Language (1815). Over the years, many other works followed, compiled also by Catholic missionaries or lay scholars, and dedicated to the analysis, translation and pedagogy of Cantonese (Lee 2005; Yue-Hashimoto 2004; Paternicò 2017). This paper will firstly offer an overview of the very early impressions and analyses of the language carried out by Westerners (given the almost total lack of previous local linguistic tools) mainly in grammars but also in the introductions of dictionaries, in manuals etc.¹; it will then show how these analyses evolved, becoming more and more refined and detailed, and eventually − thanks to the accurate work of T. O’Melia − grew more conscious and independent from the former and contemporary analyses of Mandarin, contributing ante litteram at debunking the myth of “one universal Chinese grammar”.

1 First Western Descriptions of Cantonese

The fact that Cantonese can be considered a language more than a dialect is quite accepted nowadays. Several works (Yue-Hashimoto 1993; Lucas and Xie 1994; Peyraube 1996; Matthews 1999; Cheng and Sybesma 2005), have proved that Cantonese has grammatical and syntactic peculiarities which are quite different from Mandarin, cannot always be explained as an internal development of a common language, but rather derive from the contact with non-sinitic languages. The main points raised to prove this are: the position of the modifier, which in Cantonese can sometimes follow the modified element; the position of some adverbs, which in Cantonese can appear after the verb or at the end of a sentence; the inverted order of the objects in the ditransitive construction; the way to express comparison, through gwo 過 after the adjective; the different and wider use of classifiers; the expression of potentials through V+ dāk 得.

In the very first descriptions of Cantonese, Western scholars tried first of all to retrace the origins of this “dialect”, to point out its importance, also highlighting the necessity of learning tools, given the scarcity and inadequacy of the materials locally produced. In these works, the authors acknowledged the existence of “one Chinese language”; the “dialects” being perceived as just spoken variants, where the same words simply had different pronunciations, or where occasionally different lexicon was used. This belief was very hard to eradicate for ages, if we think that in 1968 Chao Yuen Ren, in the Preface to his A Grammar of Spoken Chinese, was still stating:

When I translated my Cantonese Primer into Mandarin − for that was what Mandarin Primer essentially was − I was going to use the same eight pages on grammar, except for some minor points in which Cantonese and Mandarin differ. (Chao, A Grammar of Spoken Mandarin 1968, p. vii)

And later in the Introduction:

[…] it is in matters of grammar that the greatest degree of uniformity is found among all the dialects of the Chinese language. Apart from some minor divergencies, such as indirect object before direct object in the Wu dialects and Cantonese − for which Mandarin like English has the opposite order − and slight differences in the order of the negative in potential complements in some of the southern dialects […], one can say that there is practically one universal Chinese grammar. (Chao 1968, p. 13)

However, if the “literary style” might have a common grammar throughout China, the spoken languages do have very distinct features.

One of the oldest descriptions of Cantonese we can read was written by Eliah Coleman Bridgman (1801-1861), the first American Protestant missionary in China and the father of American Sinology². He was assigned to the China mission in 1829 and upon his arrival he began to study Chinese with Robert Morrison. In 1832 he started a mission press and the publication of the Chinese Repository, the main Western periodical in China at the time.

In the Introduction of his Chinese Chrestomathy in the Canton Dialect (Macao, 1841), Bridgeman writes about the existence of a primigenial Chinese language that over time had branched and diversified in locally spoken dialects³. To better explain his point of view he makes a comparison with European languages:

Used from time immemorial [...] this language has gradually branched into dialects some more and some less removed from the original tongue. These dialects differ from each other, like the cognate languages of Europe; and the principal ones, like the Italian and its kindred branches, require separate elementary treatises to facilitate their acquisition. Of such works the Chinese have a few [...]. But, excepting a small Vocabulary published by Dr. Morrison, no work of any note has yet been provided for the student in this dialect [Cantonese] − a dialect spoken by all the inhabitants of this metropolis, and by great numbers in adjacent cities and villages. Such neglect of this dialect, after more than two centuries of intercourse with people of Canton, ought not to be continued (Bridgeman 1841, p. 1).

Bridgeman was not the only one complaining about the scarcity of works to learn the language. The same concern was shared by Samuel Wells Williams (1812-1884), an American missionary as well as an official. He left for China in 1833 to take care of the typography of the American Board of Commissioners for Foreign Missions in Canton. From 1848 to 1851 he was editor of the Chinese Repository. Appointed Secretary of the American legation, he had a key role in the negotiations of the Treaty of Tianjin. In 1860, he was appointed chargé d’affaires in Beijing. Williams returned to his home country in 1877 and became the first professor of Chinese language and literature in Yale University⁴. He wrote a primer and a dictionary on Cantonese.

In the Introduction to his A Tonic Dictionary of the Chinese Language in the Canton Dialect, (Canton, 1856), he expresses his dissatisfaction towards the only linguistic tool on Cantonese compiled by the Chinese, which proves to be useless for a foreigner:

The standard of pronunciation for the Canton dialect is a small duodecimo handbook [...]. It is called Kong-ú ch’ik-tuk, Fan-wan ts’iú hòp tsáp 江湖尺牘分韻撮要合集 Letters for Travelers and a Collection of Important Characters divided by their Tones⁵. [...] This pocket dictionary is usually bound in four thin volumes and sold for twenty-five cents; it contains 7327 characters, and only 175 pages [...] which plainly shows how meagre are the definitions. In comparison with the local vocabularies used at Amoy and Fuhchau, it is very imperfect, and proves the ignorance of the compilers [...]. There is no table of initials and finals as in those vocabularies, nor any list of syllables, by combining which one can get the proper sound of a character; for he who uses it, strangely enough, is supposed to know already the sound of the character he is in search of. The unwritten sounds or colloquial words used by the people of Canton are nearly all omitted, which is one of the greatest defects in it, and renders it far less useful to the foreigner (Williams 1856, p. xi).

Several detailed, and sometimes curious, descriptions followed. One of them is authored by Wilhelm Lobscheid (1822-1893), a German Lutheran missionary sent to Hong Kong in 1848 to cooperate with Karl Gützlaff (1803-1851). He went back to Europe in 1851-52 to acquire more medical knowledge in London. Back to China in 1852, he became the first doctor of the Chinese Evangelization Society (CES). After being unjustly criticized by the CES, Lobscheid left China for the United States⁶. He authored several works on Cantonese.

In the Introduction to his English and Chinese Dictionary, with the Punti and Mandarin pronunciation (Hong Kong, 1866-1868) he tried to describe Cantonese (which he called, like others, punti 本地, meaning local language) and other dialects with interesting comparisons to European languages:

If we deduct the dialects which are only spoken by a few hundred thousand or a million of people, such as the Shanghai and Ningpo dialects, then we have:

1. The Punti, which in sound may be compared to High German⁷;

2 The Hakka, which in sound resembles the Swedish and Danish;

3 The Fukien or Hoklo, which we would call the Chinese French; and

4. The southern and northern Mandarin dialects, which, with their large number of gutturals or fricative sounds resemble more the Dutch of the Germanic language.

The Punti or Canton dialect is that spoken by the race who, according to Chinese accounts, is the result of the intermarriage of the aborigines with the invaders of their territory which is put down between BC 220-250. The dialect of this race is the best articulated of all the dialects we heard. (Lobscheid 1866-1868, p. 8).

It must be admitted that western scholars did not immediately acknowledge the peculiarities and the distinctive traits of Cantonese grammatical and syntactical structures. For example, Bridgeman (1841, p. ii) simply states that: “The characteristics of the Canton dialect are limited to the pronunciation, choice, and collocation of words”, and in the paragraph devoted to the grammar (p. xv-xvi), he merely suggests the use of Joseph Henri de Prémare’s Notitia Linguae Sinicae (Prémare 1831)⁸ and summarizes some general statements that can be found in Abel Rémusat’s Eléments de la Grammaire Chinoise (Rémusat 1822), which however are both grammars of Mandarin. In the rest of his work, all the explanations are lexical or cultural. It can be assumed that, not acknowledging at first much difference between Cantonese and Mandarin, aside from those concerning lexicon and phonetics, the early Protestant missionaries in Canton and Hong Kong thought that the previous grammatical analyses produced by the Catholic missionaries and lay scholars would suffice.

In other cases, it appears that the stress on a natural way of learning the language (like a child does), avoiding an excessive focus on grammar, was more of a conscious, didactic choice⁹. Samuel Wells Williams in the preface to his primer Easy lessons in Chinese (1842, p. ii-iii), a work with progressive exercises (mainly reading and translation) to learn Cantonese, states that in his work almost all the grammatical remarks have been omitted since it would be preferable for the learner to start by simply memorizing some words and sentences: “He need not be solicitous about the rules of grammar or the elegancies of style, until he has acquired a stock of words and phrases in which, as in examples, he can instantly see the application of the former, and relish the niceties of the latter.” The same attitude can be found in a work compiled more than 70 years later by R.T. Cowles who, in the preface to his Inductive course in Cantonese (Cowles 1915, p. 2-3) writes: “It is not to be hoped or expected that the book will satisfy sinologues for it aims not at presenting the language in technical or grammatical arrangement […] The author’s thought has been simply to […] take the hand of the little child in the language”.

In the following pages, the first attempts at showing or describing Cantonese peculiar grammar will be presented.

2 First Grammatical Descriptions

2.1 Robert Morrison

The very first ‘sort of’ grammatical description of Cantonese is found in Morrison’s Grammar of the Chinese Language (1815), a long work in English with examples in Chinese characters and Romanization. Despite the focus being on Mandarin, the influence of Cantonese is evident in many of the examples. A quite long paragraph (Morrison 1815, p. 259-267) is dedicated to the “Canton Dialect”, where, after a short introduction, Morrison lists the peculiar Cantonese words and characters (when available) used for pronouns, adverbs, verbs etc., often in comparison with Mandarin. Here are the main points:

• Personal pronouns: use of kue¹⁰ 佢 (kéuih)¹¹ for the singular third person pronoun, and the use of the suffix te (deih 哋)¹² to make all the pronouns plural.

• Markers of the possessive: two options ngo tik 我的 (ngóh dīk) and ngo ka 我* (ngóh ga 㗎)

• Demonstrative pronouns: ne ko * 個 (nī 呢 go) and ko ko 個¹³個 (gó 嗰 go)

• Verb to be: hei 係 (haih)

• Adverbs of negation: im 唔 (m̀h) and mow 冇 (móuh)

Cantonese is therefore presented more in terms of substitutive words. Morrison does not really explain any of the grammatical or syntactical peculiarities of the language, however in the second part of the paragraph he provides some English sentences with Cantonese translation, from which the reader should inductively understand the differences with Mandarin. For example:

• This is better than that

  
  Ne ko how kuo ko ko

  
  (nī go hóu gwo gó go)

  
  呢個好過個個

From the above sentence the reader could derive the different rendering of the comparative sentence in Cantonese. Or:

• I’ll thank you to pass the ink to me

  
  Tsing ne kai mak kuo lei pe ngo

  
  (chéng néih hāai mahk gwo làih béi ngóh)

  
  請你揩墨過黎俾我

From the above sentence the reader could understand the peculiar dative construction with béi in Cantonese.

2.2 Samuel Wells Williams

Williams dedicates an entire chapter of his Easy lessons in Chinese (1842, p. 123-148) − which, as said above, does not provide almost any grammatical remarks − to the classifiers. He feels the need to interrupt the sequence of his exercises on reading and conversation to list the most common of them with plenty of examples. He begins stating that the previous way of referring to this peculiar class of words as numerals is not correct and creates confusion. He therefore suggests the use of classitives or classifiers. Williams does not directly explain when a classifier is used (for example, after a numeral to express quantity, after a demonstrative or interrogative pronoun): he mainly writes which classifier should be used with which noun. The rest, again, is to be understood through the many examples:

• There are ten and more bells in the temple

  
  Miu noi yau shap ki ko chung

  
  (míu noih yáuh sahp géi go jūng)

  
  (廟內有十幾個鐘)¹⁴

  
  (Williams 1842, p. 25)
  
  

• One hundred cash for ten hen’s eggs

  
  Yat pak ts’in shap chik kai tan

  
  (yāt baak chín sahp jek gāi dáan)

  
  (一百錢十隻雞蛋)

  
  (Williams 1842, p. 26)

However, no example can be found on the peculiar way of using the classifiers in Cantonese, i.e. not preceded by anything else, with determinative/demonstrative function (the so-called ‘bare classifier’, Matthews and Yip 2011II ed., p. 105-107; Chan 1999 ; Lam-Paternicò 2017, p. 134).

2.3 Wilhelm Lobscheid

Lobscheid was the very first compiler of a grammar of Cantonese: Grammar of the Chinese Language (Hong Kong 1864), in two volumes of XXXVII + 111pp. the first, and VI + 178pp. the second¹⁵.

In vol. 1, the text is opened by a preface followed by a long introduction on Chinese language and its dialects. After a list of radicals, an interesting paragraph is devoted to verbs plus resultative or directional, which are explained as important verb compounds − where the first part would be the root and the second part the termination − usually not included in dictionaries.

The grammar itself begins with explanations on the tones and the phonetics in general, and then continues with the descriptions of all the parts of the speech: classifiers, articles, nouns, adjectives, pronouns, numerals etc. Interestingly, for each class of words, Lobscheid provides plenty of examples in both − what he calls − “Literary style” and “Punti colloquial”, in English, Chinese characters and transcription. The transcription always follows the pronunciation of Cantonese. The explanations are very meager and, again, most of the “rules” must be inductively understood.

In vol. 2, the author’s aim is to present a treatise on the syntax of the written style analyzing sentences quoted from the Classics and other works (Lobscheid 1864, p. iv). After a short introduction, the text is divided into two parts: part 1. explanations on the syntax of the written style, construction of sentences and word position, the cases, proper names and appellatives, notes on the Canton Colloquial (p. 61); part 2. a variety of reading lessons closed by three pages of riddles.

The examples and the texts are in English with Chinese characters and Romanization following the Cantonese pronunciation. The text was groundbreaking and became a key reference for many learners of Cantonese at the time.

As a complete analysis of this work would deserve a separate treatise, here only some key points will be highlighted concerning Lobscheid’s understanding of some Cantonese peculiar grammatical features like the use of the bare classifier, the ditransitive constructions, the comparatives, the use of dāk 得.

Concerning the classifiers, Lobscheid, like others, devotes plenty of space and examples to each of them, explaining their usage with the right category of nouns. The examples are also set into sentences. Unfortunately, only examples of classifiers preceded by numerals and rarely by pronouns are provided, no bare classifiers appear. However, in the following paragraph entitled “The Article”, Lobscheid, providing ways to express something which actually does not exist in both ‘book language’ and ‘punti’, for the first time uses a bare classifier in the example:

• The woman lost her husband when still young

  
  Ko nü yan shéung hau shang shat liù k’ü ké lò kung

  
  (Go néuih yàhn seuhn hauh sāang sāt líuh kéuih ge lóuh gūng)

  
  個女人尚後生失了¹⁶佢嘅老公
  

  (Lobscheid 1864, I, p. 18)

In another section devoted to the articles but in vol. 2, Lobscheid offers another example of bare classifier when a noun is preceded by a relative clause with the omission of ge 嘅:

• The man who went there

  
  Hü pi ko yan

  
  (heui béi go yàhn)

  
  去彼個人

  
  (Lobscheid 1864, II, p. 61)

It must be concluded that the possibility of such a use of classifiers in Cantonese was still somehow unconscious.

The same unconsciousness, but at a lower level, appears in the use of the ditransitive constructions. Although here and there we can find examples, no specific paragraph is devoted to this peculiar feature. However, in the section actually describing “The Indefinite article”, where two sentences are offered to counterpose the book style and the Cantonese colloquial, the ditransitive construction is used:

• I gave him a pear

1. Ngo pí t’á yat ko shà lí

   
   (ngóh béi tā yāt go sā léi)

   
   我俾他一箇¹⁷沙梨

2. Ngo pí yat ko shà lí kwo k’ü

   
   (ngóh béi yāt go sā léi gwo kéuih)

   
   我俾一箇沙梨過佢

   
   (Lobscheid 1864, I, p. 18)

As for the comparatives, a quite lengthy paragraph is dedicated to them with plenty of examples though, as said, with little explanations:

• You are taller than he

  
  Ní pí tá kò

  
  (néih béi tā gōu)

  
  你比他高

  
  (Lobscheid 1864, I, p. 25)

but in Cantonese colloquial:

• He is better than I

  
  Kü hai hò kwo ngo

  
  (kéuih haih hóu gwo ngóh)

  
  佢係好過我

  
  (Lobscheid 1864, I, p. 27)

Though listing several examples, Lobscheid does not seem to distinguish very well the different usages of dāk 得, somehow mixing its potential and adverbial functions (Matthew and Yip 2011, p. 203-205; 277-280). Concerning the peculiar Cantonese expression of potential through the use of verb + 得, examples can be found in the section devoted to the “Auxiliaries”. Again, book style and colloquial are compared:

• He cannot come

1. Loi pat tak

   
   (lòih bāt dāk)

   
   來不得
   

2. M loi tak

   
   (m̀h lòih dāk)

   
   唔來得

   
   (Lobscheid 1864, I, p. 60)

A few examples in the same sections actually show the adverbial use of dāk but are wrongly associated and translated as potentials, e.g.:

• *Cannot write quickly¹⁸

1. Sé tak pat fái

   
   (sé dāk bāt faai)

   
   寫得不快
   

2. M sé tak fái

   
   (m̀h sé dāk faai)

   
   唔寫得快¹⁹

   
   (Lobscheid 1864, I, p. 60)

2.4 Benjamín Castañeda

For the sake of completion on this early stage, a few lines will be added on the Cantonese grammar of an almost unknown author: Benjamín Castañeda (1846-1913). Peruvian musician and piano teacher, Castañeda spent a few years travelling the East and dedicated some publications to China and Japan. Among them, the Gramatica Elemental de la Lengua China Dialecto Cantones (Hong Kong 1869) ²⁰. This work, in 137pp., is opened by a general introduction on the language, the characters, the phonetics and the radicals (where Castañeda refers to Williams). The following chapters are devoted to the parts of the speech (articles, nouns, adjectives, measure words, verbs etc.). The explanations in Spanish are quite short but the examples are abundant. After a section on the written language, poetry and Classics, the text is closed by translation exercises of passages taken from The Great Learning and The Doctrine of the Mean.

The work unfortunately is not innovative and does not represent a relevant advancement in the field. Castañeda’s explanations appear to be quite influenced by the grammatical explanations of Mandarin and he does not appear to catch the peculiarity of Cantonese in the use of classifiers, adverbs, modifiers etc. This could be explained by his relatively short stay in the East and by his underestimation of the difficulties of the language, since in the Preface he states:

Debo añadir aquí, que el estudio de la lengua China no presenta esa gran dificultad que se la ha querido siempre atribuir. Muy al contrario, la construcción gramatical de este idioma, es quizás una de las mas sencillas que se conocen. […] La Gramática China propiamente dicha es poco complicada; las partículas que forman la clave principal del idioma, equivalen á las terminaciones del Latin; la posición de las palabras determina por si sola el valor de cada una y esto bajo reglas tan precisas y constantes, que no cabe jamás la meno incertidumbre en el sentido de una frase, y aunque la lengua sea elíptica y á menudo figurada, no puede en ningún caso una oración completa ser interpretada de dos modos. (Castañeda 1869, p. 2-3)²¹

3 Deeper Insights into Cantonese Grammar

3.1 Nicholas Belfield Dennys

Nicholas Belfield Dennys (1839-1900) was a British officer who joined the Consular Service in China in 1863 as a student interpreter at Peking. However, one year later he resigned and became proprietor and editor of the China Mail, Hong Kong, retaining the position until 1876, while he was also serving as Secretary of the City Hall and Secretary of the Chamber of Commerce. In April 1877, he was appointed Assistant-Protector of Chinese in Singapore and Justice of Peace for the Straits Settlements. Dennys was the author of several books, including: Notes for Tourists in the North of China (1866), The Treaty Ports of China (1867), The Folklore of China (1876), A Descriptive Dictionary of British Malaya (1894), and of A Handbook of the Canton Vernacular of the Chinese Language (1874) (Wright and Cartwright 1908, p. 750).

Dennys proved to have quite a thorough understanding of the language. He was the first to dispel the myth that spoken Cantonese if written down becomes understandable to all the Chinese (Dennys 1874, p. III). His handbook contains a series of lessons with some vocabulary, a few − at time specific − grammatical explanations and plenty of examples. He does not add too much to what was already understood by Lobscheid, but his explanations are more detailed and the peculiarities of the language a touch more “conscious”.

For example, concerning the “Conjunctions” he writes:

In ordinary colloquial, the copulative conjunction is often replaced by the verb when several nouns are coupled as, what we should term, nominatives or accusatives of a verb. Thus: give me tea, sugar and milk maybe translated give tea, give sugar, give milk me. (Dennys 1874, p. 4)

He then offers several model sentences like:

• Give me tea (and) sugar

  
  pi ch’a, pi pak-t’ong ngo

  
  (béi chàh, béi baahk tòhng ngóh)

  
  俾茶俾白糖我

  
  (Dennys 1874, p. 4)

Regarding the comparatives, the examples are very similar, if not identical, to those in Lobscheid (e.g. n. 9) but fundamental explanations on the word order (where to put for example 過 or 更) are provided:

The ordinary comparison of adjectives in the Canton vernacular is very simple, the word kang 更 for the comparative, and chi 至, ting 頂, kik 極 or tai yat 第一 for the superlative, being prefixed to the word; e.g. good ho 好, better kang ho 更好, best chi ho 至好.

Another way of forming the comparative is to use kwo after the adjective; e.g. He is better than I, K’ū hai ho kwo ngo 佢係好過我; This is better than that, ni-ko ho kwo ko-ko 呢個好過嗰個.

The comparative is also formed by placing kwo ü 過於 after the adjective, e.g. Older than I, tai kwo ü ngo 大過於我. (Dennys 1874, p. 81)

Thus, with Dennys’ Handbook the learner is not left alone anymore in interpreting the grammar rules or constructions inductively through a series of pattern sentences. This trend will be continued and further developed by J. Dyer Ball.

3.2 James Dyer Ball

Dyer Ball (1847-1919) was a sinologist born in China from a missionary. He worked for the Hong Kong Civil Services for 35 years holding different positions, from security officer to interpreter for the Supreme Court (1878). He was editor of the China Review from 1900 to 1901. He was a very prolific writer and reached a certain celebrity for his encyclopedic work Things Chinese: Being Notes on Various Subjects Connected with China (I ed. London 1892, IV ed. New York 1904).

He dedicated several works to Cantonese²². The first and most famous is Cantonese Made Easy (Hong Kong, I ed. 1883, IV ed. 1924) in 200pp. It was followed by: Cantonese Made Easy Vocabulary (Hong Kong, I ed. 1886, III ed. 1908), a short vocabulary for the manual; An English-Cantonese Pocket Vocabulary, Hong Kong 1886 (II augmented ed. 1894) of just 23pp.; How to Speak Cantonese (Hong Kong, I ed. 1889, II ed. 1902) of 229pp., which is a collection of 50 conversations on daily topics (like: Something to do, Something to drink, The Rain, My mother is ill, Oh! Those Lawyers, Jewelry, Fruits, The Typhoon, Military, Diplomatic Government, Exports and Imports etc.) for missionaries, merchants but not exclusively; Readings in Cantonese Colloquial (Hong Kong 1894) in 171pp., including 30 selected readings mainly chosen among translations of holy texts but also local literature.

Dyer Ball’s grammar considerations are found in Cantonese Made Easy, a primer including: a very insightful introduction to Cantonese, to its sound system (and transcription), to the tones and the way to mark them, plus drills to learn them; a phrasebook ordered by topic, with both free and literal translations; finally, a short treatise on Cantonese grammar.

In the preface the author highlights the importance of the “correct pronunciation of pure Cantonese” and refers to the syllabary published by Parker in 1880 for the orthography with some adjustments (Parker, 1880, p. 363-382). Dyer Ball is the first scholar to stress the importance of referring to Cantonese as a “language” instead of a “dialect”:

Cantonese has its ‘real dialects’ some of which are spoken by tens of thousands, or hundreds of thousands of natives, and which, if they were spoken by the inhabitants of some insignificant group of islands in the Pacific with only a tithe of the population, would be honored by the name of languages (Dyer Ball 1888, p. XIV).

In the analysis of his work I will refer to the II edition (1888), where the grammatical part was enlarged with 36 additional pages.

The author’s attention to grammar, as stated in the introduction, explicitly aims to prevent the learner from making tremendous mistakes, and at the same time at proving that the idea that Chinese has no grammar is wrong (Dyer Ball 1888, p. v-vi). As a matter of fact, his explanations are very long and detailed.

For example, he makes an unprecedented description of classifiers. In the section devoted to them (Dyer Ball 1888, p. 44-59), Dyer Ball explains the importance, variety and usage. He explains the different positions a classifier might have: after a numeral and before the noun, but sometimes after the noun; after an adjective or adverb+adjective, when there is no numeral; after a demonstrative pronoun; between two adjectives. However, he fails to include the possibility of simply having a classifier preceding a noun. A bare classifier appears in the section devoted to the “Article” where he shows the way of rendering the indefinite and definite articles:

• 一個人

  
  Yat ko yan

  
  (yāt go yàhn)

  
  man

• 個人

  
  Go yan

  
  (go yàhn)

  
  The man

  
  (Dyer Ball 1888, p. 43)

As for the comparatives, he writes about the use of the adverbs kang 更 (gāng) or chung 重 (chùhng) before the adjective (followed or not by dī 啲), ex. kang tai ti 更大啲 (gāng daaih dī) ‘larger’. He also adds that, when a term of comparison is added, the adjective is not accompanied by any sign, but the construction becomes:

• 乾過嗰個
  

  Kon kwo ko ko

  
  (gōn gwo gó go)

  
  Dryer than that
  

  (Dyer Ball 1888, p. 61)

Dyer Ball also for the first time presents the peculiar Cantonese way of rendering the comparative of equality:

• 個隻咁大

  
  ko chek kom tai

  
  (go jek gám daaih)

  
  As large as that one

• 好似呢條咁長

  
  ho tsz ni tiu kom chöng

  
  (hóu chíh nī tìuh gám chèuhng)

  
  As long as this one (lit. like this one so long)

  
  (Dyer Ball 1888, p. 61)

Dyer Ball makes a very clear explanation of the ditransitive construction stating that the Indirect Object follows the Direct Object when preceded by kwo 過, but it may either precede or follow when kwo is not used, although it usually follows:

• 俾個部書過我

  
  Pei ko po shū kwo ngo

  
  (béi go bouh syū gwo ngóh)

  
  Give that book to me

• 俾部書佢
  

  Pei po shü k’ui

  
  (béi bouh syū kéuih)

  
  Give him a book

  
  (Dyer Ball 1888, p. 90)

The potential and adverbial usages of dāk are still not very well distinguished in Dyer Ball who mixes them in the section of the “Adverbs of Manner” (Dyer Ball 1888, p. 104), stating that they can take different positions in a sentence with a slight change of meaning and then listing:

• 佢寫得快 k’ui se tak fai, (kéuih sé dāk faai) *he can write quickly²³

  
  佢寫快都得 k’ui se fai to tak, (kéuih sé faai dōu dāk) he can write quickly

  
  佢快都寫得 k’ui fai to se tak, (kéuih faai dōu sé dāk) he can write quickly

  
  快都寫得嘅 fai to se tak ke, (faai dou sé dāk ge) it can be written quickly

  
  (Dyer Ball 1888, p. 104)

All in all, Dyer Ball’s understanding of Cantonese was very deep, his analyses accurate and detailed. He shows a thorough knowledge of all the works compiled on the language before and during his time, and scientifically refers to them.

4 Setting the rules and paving the path to an independent grammar

At the beginning of the 20th century, several works were compiled for the pedagogy of Cantonese, however, none of those I could consult seem to have contributed in pushing the grammar analysis any further²⁴. We would have to wait until Thomas O’Melia (1898-1973)’s First Year Cantonese (Hong Kong 1938)²⁵ to see Cantonese grammatical consciousness moving to a higher level and detaching from the description of Mandarin (Lee 2005, p. 76).

4.1 Thomas O’Melia

O’Melia entered the American International Catholic Mission Movement of Maryknoll in 1914. He was assigned to the China missions in 1923 and was appointed director of a language school that the Maryknoll fathers had just opened in Hong Kong in 1934. He moved to the US during World War II, but went back to Hong Kong between 1959 and 1967, again as Director of Language Studies in Stanley. He moved to Chicago in 1968 to serve with the Chinese Missions and died in Philadelphia five years later.²⁶

O’Melia’s First Year Cantonese immediately became very popular. It was used by the Hong Kong Government as the syllabus for the Final Examination in Cantonese, organized by the Board of Examiners for the so-called Cadet System (modelled on the Chinese imperial examinations system), aiming at the recruitment of officials to work in Hong Kong with competence in spoken and written Cantonese (Lee 2005, p. 42). The work is divided into four parts: the first aims at ensuring a proper command of grammar “illustrating the principles of sentence structure”; the second part aims at the development of a given topic through easy compositions, using the studied grammar and sentences in context and acquiring new vocabulary; the third, presenting actual conversations between Chinese, aims at reviewing what was studied in the previous parts, “cultivating a smoother control and more fluent use of ordinary colloquial”; the forth containing all the words and phrases of the preceding parts, arranged alphabetically and annotated.

The focus of this work, on the grammar awareness of the adult learner, is very well presented in the preface written by the Superior of Maryknoll, James E. Walsh:

The method used in this manual is in principle that of apperception, and this would seem the reasonable approach for the developed mind already encrusted with grammatical habits. It attempts to build up along the lines of association and least resistance, rather than to tear down until the tabula rasa of kindergarten is reached. The child’s random way of learning a language is perhaps the best way for a child, but the process demands the wax-like mind, sharp ear, and flexible tongue of the child. It is sometimes said that the first step in learning a new language is to forget the old one. If taken literally, this would be putting oneself in the position of a child indeed, but without the child’s special paraphernalia for learning languages […] The method of apperception takes the grown man as he is, with all his blushing grammar thick upon him, and helps him to go forward along the lines of that accustomed framework. Changes and adjustments there will be. But a pleasant mastery will grow from the intelligent comparison involved. (O’Melia 1938, p. III)

The following analysis will concentrate on the first volume, where most of the grammatical descriptions are concentrated. Each of the 42 lessons is introduced by a growing number of new words, followed by several pattern sentences; a section of “Grammatical Remarks” is last, where grammar is often presented with a set of “rules”. This trait, not simply being that of an “old fashioned” descriptive grammar, is actually very revealing of the high status O’Melia confers to Cantonese language since, up to recent times it was thought to have no relevant grammar peculiarities²⁷. Already at the end of Lesson 1, O’Melia, pointing out the importance of word order in Chinese, declares what is to come: “When we wish to construct a sentence we must follow the rules of Chinese word order, under pain of not being understood. What the rules of Chinese word order are we shall learn, one by one, as we procede through the Lessons.” (O’Melia 1938, p. 3)

Each Lesson is dedicated to a particular grammar topic, with plenty of examples and detailed grammatical explanations. The peculiarities of Cantonese language are all stressed. For example, when talking about the “rule” that modifiers precede what they modify, he does not fail to inform the learner that he will find consistent variations from this general rule. (O’Melia 1938, p. 6, 18)

O’Melia devotes one of the earliest chapters to the ditransitive constructions (Lesson 4). After stating the “rule”: the indirect object follows the direct object, he identifies which verbs can be followed by two objects and in which order; the exceptions to the rule (when the indirect object is shorter it might precede), the use of bei and gwo which “govern” the indirect object but are usually omitted etc. Here are some examples he quotes:

• 我俾錢你

  
  Ngoh pei ts’in nei

  
  (ngóh béi chin néih)
  give you money

  
  (O’Melia 1938, p. 4)

• 我送花過佢
  

  Ngoh sung fa kwoh k’ui

  
  (ngóh sung fā gwo kéuih)

  
  make a gift of flowers to him

• 你送書俾佢
  

  Nei sung shuo pei k’ui

  
  (néih sung syū béi kéuih)

  
  You present him with books

  
  (O’Melia 1938, p. 5)

As for the comparatives, the explanations, patterns and examples are abundant, here just a couple:

• 長衫好睇過短衫

  
  Ch’eung shaam ho t’ai kwoh tuen shaam

  
  (chèuhng sāam hóu tái gwo dyún sāam)
  Gowns are better looking than jackets

  
  (O’Melia 1938, p. 42)

• 短衫同長衫咁好

  
  Tuen shaam t’ung ch’eung shaam kom ho

  
  (dyún sāam tùhng chèuhng sāam gam hóu)

  
  Jackets are as good as gowns.

  
  (O’Melia 1938, p. 50)

also concerning the “degree of difference”, for which he provides two examples that are both correct:

• A高兩寸過B

  
  A ko leung t’uen kwoh B

  
  (A gōu léuhng chyun gwo B)

  
  A is taller by two inches than B

• A係高過B兩寸

  
  A hai ko kwoh B leung ts’uen

  
  (A haih gōu gwo B léuhng chyun)

  
  A is taller than B by two inches

  
  (O’Melia 1938, p. 44)

O’Melia’s description of the classifiers (O’Melia 1938, p. 59-61, 70-71) is quite thorough, both in terms of position and usage, in listing the nouns that do not need a classifier, in specifying the role of dī 啲, the occasions in which you can find a classifier without noun (with noun omitted) etc. The examples are numerous and include the “bare classifier” which is however not object of any particular remark. Here are some of them:

• 啲飯好食

  
  Ti fan ho shik

  
  (dī faahn hóu sihk)

  
  The rice is good to eat

  
  (O’Melia 1938, p. 57)

• 俾啲茶我

  
  Pei ti ch’a ngoh

  
  (béi dī chàh ngóh)

  
  Give me some tea

  
  (O’Melia 1938, p. 56)

In these cases, O’Melia explains that dī is used for names which would not have a classifier. It is a quite interesting explanation − though, of course, not complete − of one of the possible usages of dī However, it must be added that the nouns he presents in the examples are indefinite or used in a partitive way in that context. They can have a classifier in other contexts. This peculiar usage of dī stresses another difference with Mandarin, where the partitive/uncountable noun would be left without measure word.

Concerning the adverbs and their position, O’Melia sets the general “rule” by which the adverbs in general precede the predicate. However, he warns about some adverbs which have a different behavior and are explained singularly. This is the case of sīn 先 which might precede the verb or be placed at the end of the sentence. O’Melia provides two examples:

• 佢打我先

  
  K’üi ta ngoh sin

  
  (kéuih dá ngóh sīn)

  
  He struck me first

• 佢先打我

  
  K’üi sin ta ngoh

  
  (kéuih sīn dá ngóh)

  
  He first struck me

  
  (O’Melia 1938, p. 96)

Then explains that in the first case the meaning would be: “He was the first to do the striking, he hit me first. (Then I struck him back etc.)”, while in the second case: “He first hit me (Then later he kicked me or whatnot)”. In a unit that is not expressively talking about adverbs, another adverb placed at the end of a sentence can be found, tīm 添 (ngoh yao m chi tim yeung tuk faat t’im 我又唔知点樣讀法添 I am further ignorant also as to how to pronounce it) but no explanations are given (O’Melia 1938, p. 251)

As for dāk, several examples show O’Melia’s (quite) correct understanding of its dual use, adverbial and potential. For example, the adverbial use is mainly (but not exclusively) treated in Lesson 20 dedicated to “Adverbs of Manner”:

• 佢行得好快

  
  k’ui tuk tak ho faai

  
  (kéuih hàahng dāk hóu faai)
  He can (or, does)²⁸ walk very rapidly.
  (O’Melia 1938, p. 105)

• 從前我讀書, 讀得好快。而家我讀書, 讀得好慢。

  
  Ts’ung ts’in ngoh tuk shue, tuk tak ho faai. I ka ngoh tuk shue, tuk tak ho maan.

  
  (chùhng chìhn ngóh duhk-syū, duhk dāk hóu faai. yìhgā ngóh duhk-syū, duhk dāk hóu maahn.)

  
  When I studied formerly I *could²⁹ study very quickly. When I study now, I study very slowly.

  
  (O’Melia 1938, p. 251)

A clearer understanding is shown when it comes to the potential dāk and its negative forms, as in the following examples:

• 我唔開得

  
  Ngoh m hoi tak

  
  (ngóh m̀h hōi dāk)
  I cannot open it.

  
  (O’Melia 1938, p. 298)

• 佢去唔去, 唔話得埋

  
  K’ui hui m hui, m wa tak maai

  
  (kéuih heui m̀h heui, m̀h wá dāk màaih)

  
  It is impossible to say whether he is going or not.

• 我去唔去, 話唔埋

  
  Ngoh hui m hui, wa m maai

  
  (ngóh heui m̀h heui, wá m̀h màaih)

  
  I cannot say conclusively whether I am going or not.

  
  (O’Melia 1939, p. 303)

All in all, no one can doubt that O’Melia’s work, with such a thorough understanding of the peculiarities of the language, paved the path for modern grammatical descriptions of Cantonese. As a matter of fact, it is the oldest grammar appearing in the list of references of Matthews and Yip’s Cantonese. A comprehensive Grammar (I ed. 1994, II ed. 2011), at present the most modern Cantonese grammar written in English, which has the merit − among the others − of having moved the analysis from prescriptive to descriptive.

Although beyond the scope of this study, it must be said that another scholar indirectly had a key role in shaping Cantonese grammar: the famous linguist Chao Yuen Ren. Indirectly because, not his Cantonese Primer (Chao 1947) − with only a short chapter devoted to grammar, plus Chao not being a native speaker −, but his pioneering work on Mandarin, A Grammar of Spoken Chinese, though minimizing the grammatical differences between dialects, had, in terms of methodology and framework, a strong influence in the compilation of Cheung (1972). Cheung’s was the first book-length grammar of Cantonese ever compiled by a native scholar and currently is the main reference in Chinese for Hong Kong Cantonese³⁰.

Conclusions

Given the lack of locally compiled linguistic tools, Western scholars were pioneers in carrying out the first analyses of Cantonese between the 19^th and the early decades of the 20th centuries. Although at first believing that Cantonese was just a branch of the Court Speech, they very soon discovered the particular traits of what deserved to be called a “language” more than a “dialect”. Their descriptions of Cantonese became more and more detailed, accurate and conscious, finally departing from the grammars of Mandarin. The apex was reached with O’Melia’s First Year Cantonese, which, although written in the old-style prescriptive manner, was compiled more than thirty years before native scholars recognized the peculiarities of Cantonese language, paving the path for modern analyses.

Primary sources

Secondary sources