Access
This article is part of Nature's premium content.
Published online 10 October 2007 | Nature | doi:10.1038/news.2007.152
News
How 'holp' became 'helped'
Studies link frequency of word use to how fast words evolve.
The less often a word is said, the faster it will change over time, whereas more-often uttered words are resistant to change. In this week’s Nature, two groups publish analyses of this trend, which quantify it and compare it with biological evolution.
To read this story in full you will need to login or make a payment (see right).
Comments
Reader comments are usually moderated after posting. If you find something offensive or inappropriate, you can speed this process by clicking 'Report this comment' (or, if that doesn't work for you, email webadmin@nature.com). For more controversial topics, we reserve the right to moderate before comments are published.
Child-language acquisition also tells us how verbs change. A copy of the above article in USA Today mentions that verbs of one syllable that begin with two consonants--the example they used was grow--change to the past tense according to the following rule: substitute ew for ow. We know children start out by saying growed before they learn the ew rule. But, what about the word snow? We don't change to past tense by saying "It snew last night." We say "It snowed last night." What does that say about the ew rule? Also, I Google a lot but always begin the word with a capital g. I have no idea why I insist that I must begin the word with a capital letter; though to do otherwise seems more like stealing someone else's word.
I am a french linguist. I would like to say to Mark Pagel that his hypothesis might be valid, but there are huge exceptions that don't fit with the hypothesis: for instance about "two", "dos", "zwei" for number 2, if the frequently used words are more conservative than the less used ones, how to explain the strong distortion of the form erku in the armenian language (another indo-european language) meaning also "two" and coming from the same root?
The poor grammar in the title and text of this article suggests a need for another study that tracks the meme of using adjectives when adverbs are appropriate. I wonder how "fast" that meme evolves.
I am a young Biology student in Mexico, and may not understand much of this and many other matters, but this theory makes a lot of sense to me. And some of the exceptions, for instance "erku", might be explained by horizontal transmition
I'm an English teacher so I often find myself cringeing at the poor usage that seems rampant these days, especially on the internet, but the reality is that living languages change. Railing against common misusage is missing the boat. To continue the analogy, you have to go with the flow. If 80% of a language's users follow some pattern of usage, that pattern is the de facto rule, however much prescriptive grammarians may dislike it. Also, finding one or two (or twenty) exceptions to the findings of the authors of these papers does nothing to discredit their work or their findings. Languages are extremely complex and the 'erku' example just shows that some other things are going on here at the same time. Not surprising.
In reply to Rick, we note that according to the Oxford English Dictionary, "fast" can indeed be legitimately used as an adverb. We admit however that in this particular sentence (the one under the headline) "fast" could be potentially be misread as an adjective applying to "words" - so "how quickly words evolve" or "how rapidly..." would have been more grammatically precise.
Sorry to be off topic, but Michel, did you play the drums in Brussels in the early 1980s?
As one who has done market research for several decades, this is a fascinating and revealing topic, especially as we consider how dependent we are on language (written and oral) to communicate meaning. For a thorough treatment of the topic, check out the "Story of Human Language" by linguist John McWhorter, PhD. His 36-lecture series addresses many forces that drive changes in language, among them biological, historical and socio-political forces. Follow this link for a bio and course information. http://www.teach12.com/ttcx/coursedesclong2.aspx?cid=1600&pc=Professor
If we believe humans are intelligent. Then I think the study of how languages change over time is better modeled by the intelligent design theory and creation than by evolutionary theory. Languages were created by intelligent beings that desired to communicate with each other. They have with thought and consideration of outcome decided to choose which words best facilitate the communication, which achieve the desired end result. Human communication is a complex process which requires intelligent modification and adaptation. As best as we know the cows moo the same way all over the world and have since humans have known them. If you don’t believe humans are intelligent then the whole discussion is meaningless noise.
Nicola, Thanks for the reply. I stand corrected on the dual role of "fast" (which is perhaps a meme itself) and didn't even notice the lexical ambiguity. Dale, I don't see how language is any better modeled with a creationist framework than is any other complex behavior. I also don't know what that framework would look like and have never seen it articulated. The only model that comes to mind is saying that language is complex because God made it that way. But there's nothing testible there and I don't see how that could better explain the way that irregular verbs become regular over time.
Interesting article, but the author fails to note that the change from a strong verb (one that changes the vowel to indicate tense) to a weak verb (one that adds "-ed") is not entirely a one-way street. Note how "dove", "snuck", and "drug" have replaced earlier "dived", "sneaked", and "dragged", at least for some speakers, and how "hung" has largely ousted "hanged" in the sense 'executed by hanging'. For Michel, note that, even though it looks really weird, Armenian /erk/ is the _normal_ outcome of Proto-Indo-European word-initial *dw. "Erku" has simply changed in the same (admittedly odd) ways as other Armenian words. Actually, English "two" is more irregular than the Armenian form: according to the sound changes that have affected English, it should be pronounced /two/ ("twoh"), rather than /tu/ ("too"). Similarly for English "one", which should be pronounced like "own" but instead is pronounced like "won". At least in the Indo-European languages I'm familiar with, numerals are kind of an exception to the "common = resistant to change" rule, in that they display all kinds of irregularities and analogical formulations.
Dale - As an experienced software engineer, I have found that compex systems sometimes abide by amazingly simple rules. This is just one example of that.
I am a linguist from Russia. Interesting idea. I think that 'evolutionary biology backgrounds' are explained by simple illiteracy of people and unwillingness to learn one's own language.
don't get me started on intelligent design and creationist theory dale.... science and religous are two different things one will use evidence and can be dissproven whilst another is based upon a story. maybe you should read the god delusion by Richard Dawkin
Answer to Mrs. Stocki: Considering the "Many Universes" theory science seems not to use evidence some times. Another "story" is the second law of thermodynamics which requires a closed system that no one can observe and less making it evident to me. For sure pure randomly given things will as likely prepare an incompressible coding DNA set as observing transfer of all heat to a single gas molecule in your office. Conclusion: Rules for evolving information process elements are very important, may be there is some general rule behind.
Language also can evolve due to communication from a person with a different native language. This kind of evolution is similar to child-language evolution. Even words from other languages can be adapted phonetically into another language, for example the word 'catamaran' meaning a small wooden raft, in tamil and english, can be considered as another evolution.
I really liked the Lewis Carroll reference. Carroll did such clever things with math and words in his work, so it seems appropriate that his Cheshire cat makes a disappearance.
It is common knowledge among linguists that lexical change is, generally speaking, inversely proportional to frequency of use. It would be incorrect, however, to claim that frequency of use is the *only* factor motivating or inhibiting change. Contrast, for example, auxiliary verbs in French (« avoir » and « être ») with some other first conjugation verbs, like « manger ». As auxiliary verbs are used in all of the composed tenses, they are very high-frequency and have thus undergone relatively little change over the years when compared with other verbs. Now consider the (suppletive) forms of "goose" and "geese" which should have extremely low usage compared with other lexemes which are high-usage and demonstrate strong resistance to change. The paper is very interesting; I'm very glad to see more empirical linguistics appearing in Nature!