Hyphenation, or how I survived the 2009 spelling reform

As I mentioned in my earlier post entitled The Spelling Revolt, the Portuguese citizens’ manifesto decrying the 2009 spelling reform also strongly criticizes the new rules on hyphenation contained in the reform which it describes as ‘totally chaotic.’ Here I give my view on the matter and recount my own experience of the spelling reform.

Hyphens have a very important role to play in Portuguese: they are used to append unstressed object pronouns to the verbs they are associated with when the syntax of the sentence requires the pronoun to be placed after the verb form (e.g. sentem-se, por favor). In Brazilian Portuguese, this is something that mostly occurs in more formal and literary style, except in the case of the special forms of the third person pronouns when appended to infinitives (-lo, -la, -los, -las) which are also used in neutral and more informal registers. But in addition to this most important function, hyphens are also used after prefixes in certain circumstances (e.g. recém-nascido, micro-ondas etc.) as well as to link individual words that go together to form a single semantic unit (e.g. grão-de-bico ‘chickpeas’, bom-senso ‘common sense’ etc.). The use of hyphens in these last two cases is subject to a set of hard and fast rules laid down by the Brazilian Academy of Letters in Brazil and the Lisbon Academy of Sciences in Portugal. These rules must be memorized by millions of school students and candidates in public service selection tests. Changing the rules, as happened in the 2009 spelling reform, would only be desirable if it were to make them simpler and easier to implement, but unfortunately that was not the case.

The manifesto cites a few of the resulting inconsistencies, such as guarda-chuva vs. mandachuva, but there are many more. For example, why did the word paraquedas lose its hyphen in the reform, but para-brisa, para-choque and para-lama didn’t? The reform did away with hyphens in three-word compounds except those used as names of plants and animals. So bico de papagaio referring to the medical condition known as a bone spur is now spelt without hyphens, while bico-de-papagaio, the Portuguese name for the plant poinsettia, retains them. This rule is reasonably logical and would be easy enough to apply were it not for the fact that the reform allows for “exceptions consecrated by usage, such as água-de-colônia, arco-da-velha, cor-de-rosa, mais-que-perfeito, pé-de-meia, ao deus-dará, à queima-roupa.” Isn’t that a highly subjective criterion? And weren’t all three-word compounds previously spelt with hyphens “consecrated by usage”? It’s also unclear in the wording of the reform whether this is a finite list of exceptions, or merely a selection, as “such as” would suggest. More to the point: if you lay down a rule and then allow exceptions based on no objective criteria, how is the language user to apply the rule with any certainty?

Some months before the 2009 reform was introduced, I was asked by the British publisher Collins to update the spelling in their range of English-Portuguese dictionaries (Gem, Escolar and Prático) which I had been in charge of creating in the late 1980s. This job turned out to be something of a nightmare as, although implementation of the reform was imminent, the Academia Brasileira de Letras had not yet published an updated version of its Vocabulário Ortográfico da Língua Portuguesa (VOLP), the reference list of how every single word in Portuguese should be spelt. (Every single word? No, of course not! It can’t possibly keep up with the rate at which new words are coined in the language, making it a rather futile undertaking.)

In Portuguese dictionaries, hyphenated compounds are given headword status, so pé-de-meia ‘nest egg’ has its own entry, separate from the entry for . Unhyphenated compounds are traditionally subentered under the first word, so pé de atleta ‘athlete’s foot’ appears in the entry for . So I had to delete the entries for compounds which were hyphenated pre-reform (e.g. pé-de-atleta, pés-de-galinha, pé-de-moleque, pé-de-pato, pé-de-vento) and incorporate the new unhyphenated compounds into the entry for the first element, namely in the examples given. This was straightforward enough, because these compounds only occur as masculine nouns, like itself, so they don’t require any further grammatical labelling in the dictionary. But then I came to the case of cara-de-pau, now to be spelt cara de pau, which can be a feminine noun meaning ‘nerve, brazenness’, but can also be used as an adjective meaning ‘brazen’ and as a two-gender noun meaning ‘a brazen person’. The feminine noun usage was spelt without hyphens pre-reform and incorporated into the entry for cara, but what to do about the other usages previously spelt with hyphens and having an entry of their own? And what about anomalous three-word compound nouns like chove não molha ‘shilly-shallying’ and deus nos acuda ‘uproar, commotion’? In desperation, I made an executive decision: to keep cara de pau and chove não molha as separate entries, even though as far as I am aware it was the first time ever that an open (unhyphenated) compound had had its own entry in a Portuguese dictionary. It’s common practice in dictionaries of English, but it went against all the traditional rules of Portuguese lexicography. It seemed acceptable to incorporate deus nos acuda into the entry for deus because it is only used as a masculine noun, as is deus itself.

I have been reassured by the fact that, since then, other dictionaries have done the same thing, although we now have the peculiar situation whereby cara de pau (and unhyphenated compounds like it) appear in the dictionary twice, once in the entry for cara, and once as a separate entry. At least one of the dictionaries I consulted while writing this gives examples using the feminine noun sense (‘nerve, brazenness’) in the wrong entry, which well illustrates the absurdity of having the same item, with the same spelling, in two separate places in the dictionary. The new hyphenation rules also meant that the medical sense of bico de papagaio would now have to be subentered under bico, while the botanical sense, still spelt bico-de-papagaio, would continue to have its own entry.

In general, I’m in favour of doing away with hyphens in compound nouns, or at least allowing users of the language to hyphenate compounds as they see fit, as we do in English. The problem with imposing rules is that, if they are not easy to apply to any kind of compound, they simply generate confusion. The spelling bible, the Vocabulário Ortográfico da Língua Portuguesa, can never hope to include all the words in the language and even though it can now be consulted for free online so you no longer have to buy the expensive book, the content hasn’t been updated since 2009. Just think of all the new compounds that have entered the language since then. The hyphenation rules state that compounds consisting of a noun and an adjective should be hyphenated if they constitute a single semantic unit. Thus, the VOLP lists caixa-preta ‘black box (of an aircraft)’ spelt with a hyphen. But following the same principle, we should also hyphenate compounds like buraco negro ‘black hole’, caixa eletrônico ‘ATM machine’, casa lotérica ‘lottery outlet’ and disco rígido ‘hard disk’, but nobody does. If you go to check in the VOLP, you won’t find these compounds, but does that mean they shouldn’t be spelt with a hyphen, or just that the VOLP hasn’t got around to including them yet? Or didn’t see fit to include them at all? Wouldn’t an easier rule be just not to use hyphens in any such compounds? After all, Spanish manages pretty well without them.