When you browse through introductions of many languages, you would come across terms such as ‘tonal language’ and ‘pitch-accent language’ used to describe some of these languages. For instance, the Chinese languages are among the most well-known tonal languages, while Japanese is amongst the most well-known pitch-accent languages. Furthermore, there are some languages that appear to straddle a sort-of grey area, such as Navajo. And so, I want to try to dissect this even more, and break down the differences between the two types of languages which can be said to exhibit some form of distinguishing pitch or tone.

Today, we will start with defining some terms one would commonly come across when they hear of this topic, and later on, we will start to break down what people mean by ‘pitch-accent’ languages. It turns out, that the more I read into the topic, the more I realise that this term is shakier than what it appears to be.

The notion of pitch-accent languages is covered by the field of word prosodic systems and typology, alongside notions like stress languages and tonal languages. Some linguists argue that this is more of a property-driven typology, as the latter two types of languages exhibit some properties that are present by the member languages. Tonal languages, for instance, may be defined as languages with word-level pitch features, something that is aptly demonstrated in languages like Cantonese, Thai, and Hmong. Stress languages, on the other hand, can be argued to be defined as languages that contain some sort of word-level metrical structure, as in, some syllables in a word may be stressed or unstressed. Hyman’s 2009 paper suggests that based on these widely agreed-upon properties, there would be two word prosodic prototypes. Thus, before going into the contentious subject called the pitch-accent system, let us take a look into stress and tone systems of word-level prosody.

Tone

In linguistics, ‘tone’ is traditionally defined as a ‘distinctive use of pitch’. This usually just extends to the word, as how pitch varies during the expression of a sentence is usually referred to as intonation. Unlike intonation, tone is used to distinguish between words or grammatical functions, in addition to paralinguistic processes that convey things like emotion and emphasis.

Tone may be used to distinguish between different lexical terms as well as playing grammatical roles. For example, Mandarin Chinese and many tonal languages of East Asia are predominantly lexical, where tones are used to distinguish between words that would otherwise sound identical with one another. However, some varieties of Chinese may also feature tonal inflections that reflect some grammatical function. This is mainly seen in how personal pronouns are inflected in varieties such as Hakka.

In Africa and perhaps the Americas, tone can carry both a lexical and grammatical function. In several introductions past, we have introduced the languages that inflect for grammatical case by tone, so I would encourage you checking it out for more details and examples of such languages.

Tones may be classified into two categories — register and contour. Register tones are level, with linguists largely agreeing on distinguishing up to 5 different levels of pitch. Contour tones, on the other hand, include rising and falling tones. These involve a shape, or a change in pitch. Mandarin Chinese is perhaps the textbook example of a language with both register and contour tones, perhaps alongside Vietnamese, while languages like Yoruba and Navajo are more cited examples of languages with register tones.

In some tonal languages of North America, such as Navajo, there exist just two phonemic tones, high and low. Navajo sets the low tone to be the default, with some syllables are marked as having the high tone. Other languages like Sekani use the high tone as the default tone instead. Japanese is perhaps one the most notable textbook examples of pitch-accent languages, featuring a “two-tone” system as well, but unlike Navajo and Sekani, Japanese is still traditionally classified as a “pitch-accent” language. So, what are the differences here? This is where we should introduce term called stress.

Stress

Sometimes, you might find this written as accent as well, but both terms essentially describe the relative emphasis of a certain syllable in a word. Words may also be stressed in a sentence, but that is beyond the scope of word-level prosody systems we are interested in today. With this emphasis, carries a certain prominence of that syllable within a word. While some might use ‘stress’ and ‘accent’ interchangeably, ‘accent’ might be used in sentence-level analysis of prominence of certain words or syllables, much like how intonation is the term used in a more sentence-level scale.

Stress may be expressed through various means. Usually, stressed syllables may be audibly more pronounced, resulting in them sounding louder compared to other syllables in the word. In some languages, stressed syllables may also feature lengthened vowels. Meanwhile, in languages like Portuguese and Russian, stressed syllables feature some vowel changes, such as the ‘o’ sound in Russian being rendered as /o/ in stressed syllables, and [ʌ] or [ə] in unstressed syllables. Furthermore, in languages like Mandarin Chinese, stress may be expressed using a wider range of pitch variation.

Additionally, languages may or may not phonemically distinguish by stress. Some languages have rather fixed stress rules, like in the Finnish language, where the first syllable of a word is always stressed. Others like Latin would be more flexible, where stress is determined by syllable weight and rime instead. Other languages like French may be considered to have no real word stress, as such stress can be argued to be attributed to prosody instead.

In languages like English, a stressed syllable position may be used to distinguished between otherwise identically-sounding words, or even word classes of words. For example, take the word pair insight and incite. Typologically speaking, these two are distinct words, but when we analyse the phonemes of both these words, we see that they are pretty much identical in pronunciation. However, in insight, the stress falls on the first syllable, while in incite, the stress falls on the second syllable. Another example may be the word record, which can function as a noun or a verb. If one wants to use it as a noun, as in a digital record or a criminal record, the first syllable of the word record is typically stressed. But if one wants to use it as a verb to record, the second syllable of the word record would be stressed. However, such stress patterns may vary by region as well.

Now that we have briefly looked at what tonal languages and stress languages essentially are, we can finally begin to tackle the identity crisis that is the ‘pitch-accent’ languages, which term has drawn the ire of typologists when trying to define a prototype for languages like Swedish, Japanese, and Basque. As we have seen, the term ‘pitch-accent’ quite literally combines certain features of both stress and tonal languages, with some languages taking a liking towards one of the two prototypes outlined here. So, is there a single ‘pitch-accent’ prototype, and how exactly might a pitch-accent language be defined or identified?

Further Reading

Hyman, L. M. (2009) ‘How (not) to do phonological typology: the case of pitch-accent’, Language Sciences, 31, pp. 213-238.

van der Hulst, H. (2011) ‘Pitch accent systems’, in Volume II. Suprasegmental and Prosodic Phonology, pp. 1-27.