Pronunciation in General Language Dictionaries

Janet DeCesaris

Pronunciation in general language dictionaries is an interesting topic. Today, people looking up words on electronic devices are able to hear a standard pronunciation of a word by clicking on an icon. My guess is that most people on devices that reproduce audio pay little attention to the written representation of pronunciation in dictionaries of English, which may use the International Phonetic Alphabet or some system of respelling. This is a far cry from earlier periods, when printed dictionaries contained extensive guides to pronunciation in the front matter. To cite just one example that everyone can consult freely on Internet Archive, the 1910 edition of Webster’s New International Dictionary of the English Language based on the International dictionary of 1890 and 1900 published by G. & C. Merriam (now Merriam-Webster) has an extensive (and very interesting in my opinion) Guide to Pronunciation that runs 39 pages long, including the 15 pages devoted to the ‘Synopsis of Words Differently Pronounced by Different Orthoëpists.’ In terms of pronunciation, current printed dictionaries of English contain a guide to the symbols used in representing pronunciation, but little more. It would be difficult to imagine a dictionary of English without information on how words are pronounced, however, because of the less-than-transparent relationship between English orthography and pronunciation.

Authors often contrast the complex case of English orthography and pronunciation with that of Spanish, which is assumed to be simple. Comments like the following are not difficult to find in the literature (both these quotes are taken from chapters in The Oxford Handbook of Lexicography (OUP, 2016, P. Durkin, ed.): “Polish, Hungarian, or Spanish, for example, have regular, predictable syllable stress and a simple one-to-one relationship between orthography and pronunciation. All that a speaker requires to pronounce every word originating in such languages accurately is knowledge of the relevant phonological rules and alphabet (including accents and diacritics). Guidance would only be needed for unusual items such as loanwords and place names.” (C. Sangster, ‘Pronouncing Dictionaries,’ pg. 292); “For many European languages, for example Czech, Finnish, Hungarian, Slovak, or Spanish, the problem [representing pronunciation in dictionaries intended for general readers] does not present itself in this form, since the vast majority of spellings are immediately pronounceable without any difficulty (although the precise realization may differ in different varieties of the language)” (G. Pointon, ‘The Representation of Pronunciation in General Dictionaries,’ p. 473). I certainly have no quibble with the observation that letters-to-sound mapping is much simpler in Spanish than in English, but I do think that the underlying premise of Spanish pronunciation being so obvious to readers of general dictionaries needs further comment. The Spanish Royal Academy (Real Academia Española) determines what constitutes standard Spanish around the world; words and senses that are included in the Academy’s dictionary (currently entitled Diccionario de la lengua española) are considered standard and eligible for use in edited publications around the Spanish-speaking world. If the Academy included pronunciation in its dictionary, its license of a certain pronunciation would likely exclude some pronunciations from the standard. It is not the case that the Academy is not concerned with orthography and pronunciation: in recent years, it has published books that discuss both at length (Ortografía de la lengua española, Espasa Libros, 2010, 864 pages; Nueva gramática de la lengua española: Fonética y fonología, Espasa Libros, 2011, 608 pages). I would argue that that the Academy believes that what is considered standard pronunciation of Spanish has too many variants to be included in a dictionary, and as a result it deals with those variants in a separate volume. In other words, if speakers of Spanish want to consult the pronunciation of a sequence of letters, then they must look the information up in the corresponding volume of the grammar, not in the dictionary.

A similar situation is true of Catalan. Catalan orthography and pronunciation are more complicated than Spanish orthography and pronunciation: there are grave and acute accents, and there are more sounds and more phonemes than in Spanish. Like English, there are often several letter sequences that have the same pronunciation (for example, the voiced palatal affricate like that in the English word ‘judge’ can be spelled with tj, tg, or dj, depending on the position within the syllable and the etymology; the sound [s] may be spelled with s, ss, c or ç). Catalan dictionaries do not include a representation of pronunciation precisely because both the Catalan language academy (the Institut d’Estudis Catalans) and commercial publishers want all Catalan speakers to feel represented by the information in general dictionaries so that they feel confident to speak the language in public, despite the very notable differences in pronunciation across dialects. For many years now, there has been some political pressure from certain conservative groups to “divide” Catalan speakers and convince speakers from Valencia, Alicante, and Castellón—who call their language valencià—that they speak a language that is different from Catalan (there is no linguistic basis to that argument). If dictionaries, as they strive to describe the language of an entire speech community, provide the most widespread pronunciation of Catalan, which is that of Catalonia (including Barcelona), then speakers of valencià and of other dialects (such as that spoken in Majorca, for example) may be led to believe that they actually do not speak Catalan, which would certainly exacerbate the already precarious social situation of the language, as it is under much pressure from Spanish. Of course, dictionaries could include several pronunciations, but in a context like that of Catalan this may not be feasible because of the resources available to dictionary-making and because a very large percentage of the vocabulary would need several pronunciations. There is a general feeling among Catalan lexicographers that providing only one pronunciation excludes many speakers whose pronunciation is, in fact, considered standard for a particular geographic area, and providing several pronunciations might result in speakers saying the language is simply too complicated (speakers already say that about spelling) and as a result resort to speaking Spanish, in public and in private.

In sum, I think the view expressed by Sangster and Pointon is only part of the story. English exhibits a particularly complex relationship between pronunciation and orthography, and general dictionaries do a good job of providing users with model pronunciations (it would be remiss of me to not mention here that the wonderfully clear pronunciation of our Newsletter’s editor, David Jost, can be heard on the American Heritage Dictionary website). Speakers expect general language dictionaries of English to provide guidance on pronunciation. I think most speakers likely to look up a word in a dictionary know that there are several standard ways to pronounce English and understand that the audio provided is that of standard speech, but there may be other pronunciations from other dialects that would also be considered standard. In other languages, however, the prescriptive nature of information in dictionaries is especially forceful, and as such the inclusion of a particular pronunciation may be problematic. Some lexicographic traditions have chosen to forsake including pronunciation in general dictionaries altogether, and that decision is not only grounded in the relative simplicity of the language’s written representation of sound.