What is a word?

 I did some crowd-sourced research on twitter. To try to find out what "a word" is. Some people gave me dictionary definitions. But that's a bit like trying to find the meaning of life by looking it up in a dictionary. You are going to be disappointed. 

The dictionary definition  A single distinct meaningful element of speech or writing, used with others (or sometimes alone) to form a sentence and typically shown with a space on either side when written or printed, tells me that "play" and "plays" are words. But are they the same word? Or different words?

So I asked a series of questions to see what we conceptualise as "a word". Of course by "we", I mean people on twitter at around 7.00 in the morning, predominantly teachers of and speakers of other languages. So that for example, a third of people think that played (simple past) and played (past participle) are different words. As the spelling, pronunciation and core meaning of the word are identical, this means our survey is going out to people who are making fine distinctions based in part on grammatical distinctions. So given a different pool of respondents, I think we would have got very different answers. But on the other hand, if I am asking "What is a word" then having a pool of people equipped with linguistic knowledge is possibly an advantage, as long as we are aware that this is baked in to the results.

And the results are: We don't know what a word is.

The biggest area of agreement was around core meaning. Where 2 words were spelled and pronounced the same, but had substantially different meanings (even if ultimately from the same root), then a large majority agreed that they were different words.

This example also has plays as different parts of speech. I have just belatedly posted a question to try to get at this. Using words with the same core meaning but one a verb one a noun. So far it is coming out strongly in favour of being different words.

It does seem that perceived meaning (rather than literal core meaning) is important. For example running water was seen as different to I went running. I know they are different parts of speech. But it came out particularly strongly as different words, compared to two different grammatical uses of the word playing.

Again we have almost a third of people who consider "I was playing" to be a different word to "I like playing". Based presumably on its different grammatical role in the sentence. But the 90% who consider "running water" to be a different word to "I went running"  shows the role of perceived meaning.

So I came up with other questions to get at the idea of whether words with the same core meaning but different forms counted as the same word. It turns out that again, we don't agree. But "regular" plurals with an s are seen largely as variations on the same word. Whereas proper good old English plurals such as children are seen by many as a different word. Children, bretheren, oxen, abominable snowmen. Good job I didn't include mouse/mice, cow/kine.

And then there are sheep.

Most people think sheep and sheep are the same word. But more people think that sheep and sheep are different words than think play and plays are different words. This may be a result of the fact that thinking about one question affects your thinking as you answer the next.

With verb forms, there were again three elements. Meaning, sound, spelling. Predictably, the more of these that were different, the more people considered it to be a different word. But the concept of "same word but with a different ending" was strongly detectable.

We were fairly evenly split on play / plays as a third person ending. And of course the nature of the survey doesn't tell us how arbitrary the choice was, with many people probably umming and ahing (how do you spell that?) about what to put. Most people think that "I fly" and "we fly" are the same word.

With changes of tense, the argument of "same word, different ending" was weaker. And not just with forms like fly/flew or for with changes in pronunciation read/read, but also for regular simple past tense ending in --ed. These were generally considered to be a different word.

You can see the whole thread here. I think broad conclusions are:

1. There is not clear agreement on what a word is.

2. Perceived meaning (even if literal root is the same) counts for a lot.

3. Regular plurals may be seen as the same word but with a different ending.

4. Changes of person in a verb are seen as the same word, but when there is a 3rd person s added, this is seen by some as making a new word. Generally where there is no change to form, a change of person of the pronoun is not seen as implying the verb is a different word.

5. Change of tense is often seen as being a different word. More so with different pronunciation or an "irregular" change. But also with regular --ed endings.

6. It turns out that actual linguistics linguists don't use the word "word". It's too vague! They (like mathematicians!) use the word lemma for a base concept that can be inflected or extrapolated upon.

Does this have any impact on how we teach languages?

Every so often our school has a round of cross-subject initiatives to try to metacognitisify the learning process for pupils. One variation of this did have every subject writing their generalised conceptualisation of the stages learning onto a laminated brain. For languages we had the stages of:

strategies for learning vocabulary

learning how to put words together to make a sentence

learning to change the endings of words

expressing yourself using the knowledge you have

Or something like that but better expressed. The idea of inflection of words and the way in which other languages may do this to a greater degree than English, is an important concept for our pupils. We come up against this when pupils use traditional dictionaries. Wrong word for saw. Can't find the word is. As always, important not to bypass these issues, but to tackle them head on.

Thank you to everyone who joined in and answered the questions! Hope you enjoyed it. Certainly made me think.