What parts of speech are the most susceptible, and the least susceptible, to linguistic change? And why?

I would think that nouns are the most susceptible, and that closed word classes, such as articles and conjunctions, are the least susceptible. For example, when was the conjunction “and” anything other than “and”?

And have there been cases of the parts of speech that are the more conservative elements suddenly changing?

Any ideas? Or maybe suggestions for where I can look for such a discussion?

  • 3
    Parts of speech are not natural objects, but rather artifactual terms used to describe language. They are more or less arbitrary and official, not to mention specific to each language. Therefore they aren't subject to anything natural like language change. It's the language that changes, however we choose to describe it. – jlawler Jul 11 '14 at 15:34
  • 1
    Well put @jlawler. – Dominik Lukes Jul 12 '14 at 1:55
  • 1
    What do you mean by 'change'? Addition/loss of members? Change of grammatical role of the category? Change in grammatical role of individual members? – Gaston Ümlaut Jul 13 '14 at 11:55
  • This seems to be more a request for information than a question with a precise answer. I think what you really need may be book like Guy Deutscher's Through the Language Glass that contains an easily accessible explanation of the factors driving language change. – user4938 Jul 28 '14 at 11:55

TLDR: Personal pronouns (I, you), interrogative pronouns (who, what), quantifiers (all, many, one) , negative particles (not) and determiners (this, that) are less likely to change over time than most nouns, adjectives and verbs.

There is a broad distinction between open-class and closed-class words. Open-class are those that easily admit new members, such as verbs, nouns and adjectives. They also consist of a large number of individual words - It's a daunting task to count all the English nouns listed in any given dictionary. Closed-class are those that do not readily admit new members, such as personal pronouns (I, you), interrogative pronouns (who, what), quantifiers (all, many, one) , negative particles (not) and determiners (this, that). They also consist of a rather small number of individual words - It's easy to count all the personal pronouns of the English language.

Most individual closed-class words also occur much more frequently in oral and written language than most individual open-class words. Now, frequency of occurrence happens to be one of the main factors influencing likelihood of historical change in linguistics (see, for example, Diessel (2007). Basically, the mental representation of a word is reinforced every time you hear or use it, and children are less likely to come up with innovations in language acquisition, if they hear a certain word very often. It is not true, as @Darkgamma stated, that

Nothing is inherently more conservative than anything else because pretty much every word is changed as time goes on.

Yes, every word is changed as time goes on (which is a big problem for linguists trying to learn anything about the prehistory of human languages), but they don't all change at the same rate. Historical linguists trying to determine how various languages are related exploit this by comparing words that are least susceptible to change - see for example the famous Swadesh list (although perceived cultural universality was an important consideration in choosing this list). Note that this list contains many personal and interrogative pronouns, negative particles and determiners. It also contains many verbs and nouns, but again these are frequent words referring to basic concepts.

So personal and interrogative pronouns, quantifiers, negative particles and determiners - as a group - are less likely to change over time than nouns, adjectives, and verbs - again, as a group. But this is not due to any inherent quality of pronouns and the like, but because of frequency of occurrence.

| improve this answer | |
  • Are you talking about the evolution of words from an older to a more recent form, or are you talking about the substitution of one item of vocabulary by another, etymologically unrelated, word? Not, I think, the same question at all. – fdb Aug 20 '14 at 16:32
  • Yes, they are not the same, but the frequency effect works in both cases. I'm not denying that there are universal sound changes, but above and beyond that, the "conserving effect of token frequency" (see the Diessel paper, p. 118) plays a role in both contact-induced language change and internal change. For example, irregular verbs/nouns are more likely to regularise if they have a low token frequency. – robert Aug 20 '14 at 16:56
  • Regular sound change is not impeded by high or low frequency. – fdb Aug 20 '14 at 18:50
  • As I said before, "I'm not denying that there are universal sound changes." – robert Aug 20 '14 at 20:46

Nothing is inherently more conservative than anything else because pretty much every word is changed as time goes on. Some words acquire irregularities, but these get reanalysed as regular patterns over time. A few examples are in "man/men" and "work/wrought".

Now, conjunctions are just regular words that change. Let's take your "and" for example. It derives from Old English "and/ond/end", which derives in turn from Proto-Germanic "*andi/*anþi" or "*undi/*unþi". German has from that source "und", Dutch has "en" and Scots has "an". Further away in the Germanic family, there's "ok/og/och" in Scandinavian, from Proto-Germanic "*auk" whence also English "eke", and Gothic had a suffix instead, "-uh". This is all the product of semantic shift (change in word meaning over time) and sound change (change of word shape over time)

| improve this answer | |

In principle, sound laws are universal within any given language. (I say “in principle” because every rule has its exceptions). This means that the same sound laws will affect all words regardless of whether they are nouns, verbs, prepositions or whatever.

| improve this answer | |
  • The question did not restrict linguistic change to sound change. It wasn't sound change that replaced Old English þēod with nation. – Nick Nicholas Jul 9 '18 at 14:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy