By Gordon Rugg
There’s a useful three way distinction in linguistics between things, concepts and words.
This article is a gentle examination of the distinction, with some thoughts about implications for human error.
Unicorns and non-unicorns
Sources for images are given at the end of this article
I have a deep, quiet love for linguistics. When I encountered the subject as an undergraduate, it was the first time that I really grasped what science was about. It was clean and elegant and it mapped on to reality in a way that opened up a whole new way of looking at the world.
One of the first things you learned on a linguistics course back in the day was the distinction between things, concepts and words. It’s simple, but it’s powerful.
Things means what you might expect; real, solid, tangible objects such as horses, or colours, such as grey, or observable actions, such as grazing.
Words are also what you might expect; names, or labels.
Concepts are more complex. They’re abstract mental entities, which don’t necessarily need to correspond with a thing.
Here’s an example.
The first image in the banner above is of a thing, which has a corresponding concept and word, namely rhinoceros.
So far, so straightforward. Where this three-way distinction starts to become interesting is when we look at cases like the two below.
Two concepts or one?
Sources for images are given at the end of this article
Both these images are of things. Both of them map onto the concept and word rhinoceros. However, there’s no inherent necessity for that to happen. The two animals in the photos are similar in some ways, but very different in others. In English, they are treated as sub-categories of the same core concept. I’ll come back onto the issue of sub-categories at the end of this article.
To clarify the issue involved, here’s another set of images. For most native speakers of English, the three things in the images below map onto three different concepts and words, namely horse, donkey and pony. However, the three things are biologically more similar to each other than the animals lumped together under the single concept rhinoceros – horses, ponies and donkeys are similar enough to be able to inter-breed. Treating them as three different concepts is very much a human, social distinction.
Horse, donkey, pony: Three different concepts
Sources for images are given at the end of this article
In summary, things exist regardless of human preferences; concepts and names, in contrast, are human inventions.
This leads to some interesting situations, such as when a concept and a corresponding word exist, even though no corresponding thing exists. The classic example is the unicorn, where the concept and the word exist, but not, alas, actual unicorns (despite what Caesar had to say on the subject).
Here’s a table that spells out the permutations, with an example of each.
The first row is straightforward; there’s a tangible object, with a corresponding concept and word.
The second row involves cases where something exists, and you have a concept for it, but you don’t have a simple name for it; usually, instead, you describe it with a phrase, such as “That feeling you get when X happens”.
The third row is all too familiar to anyone trying to learn a language by direct instruction in that language. Someone shows you an object, and tells you what it’s called, but you have no idea what the concept is.
Here’s an example.
Sources for images are given at the end of this article
The object on the left, with the radiating spokes around the central hole, is a tsuba. You now know the thing, and the word. What about the other three objects next to it? Are they tsubas, or not? They look similar in some ways, but very different in others; do they belong within the same concept in the culture where they originate, or are they something very different that just happens to look superficially similar?
Returning to the table: The fourth row is interesting, in terms of the error of reification; there’s a concept and a word, but there’s no corresponding tangible object. A classic example is the unicorn, where the concept is well defined in western culture, and the word is well established. However, that doesn’t, unfortunately, mean that unicorns exist. There’s been a lot of debate about various medical concepts in this regard; just because there’s a concept and a name, it doesn’t necessarily mean that they reflect an actual medical condition (e.g. the medical debate about whether Gerstmann syndrome really exists as a thing in its own right).
The fifth row maps onto, for instance, a newly discovered species that hasn’t yet been named.
The sixth row describes “gut feel” concepts that we can’t put into words, and that don’t actually correspond with reality.
The final row describes something that quite often turns up in historical linguistics, where we can read a word in an ancient text, but we have no idea what the word means, often because it only occurs in one place. This is a significant problem for biblical translators; for example, the biblical story of Noah’s ark says that the ark was made of “gopher wood”. This has nothing to do with the animal called a gopher; instead, it’s a transliteration of a word (“gopher”) which only occurs in that one place, and whose actual meaning has been lost. It’s some kind of wood, or adjective applied to wood, but nobody knows what it means.
Closing thoughts
Although the core distinction between things, concepts and words is simple, it leads into a lot of interesting issues. This article has skimmed the surface of some of those issues. For instance, there’s the question of how to handle cases where we’re not sure whether or not something exists. This was the case for the unicorn in the middle ages, where distinctive spiral ivory horns definitely existed, and helped to differentiate the concept of the unicorn from the concept of the rhinoceros, but where the source of those horns wasn’t definitely known for quite a while.
As for concepts such as the soul, or consciousness, much can be said, and already has been, about whether or not these are concepts and words without any corresponding thing.
There are also related literatures that I’ve left for a later article. I haven’t, for instance, gone into the issue of folk taxonomies, where there’s been some very interesting work by people like Rosch on the “core” levels at which people categorise, as opposed to the sub-type distinctions that are made using modifiers, such as “Javan rhinoceros” as opposed to “black rhinoceros”. Similarly, there’s a large and fascinating body of research into the way that language reflects what a culture cares about.
I’ll be returning to this theme in another later article, looking at how languages handle numbers, and at the implications for how most people handle estimates of probability.
Notes and links
Sources of images:
“Waterberg Nashorn2” by Ikiwaner – Own work. Licensed under GFDL 1.2 via Commons – https://commons.wikimedia.org/wiki/File:Waterberg_Nashorn2.jpg#/media/File:Waterberg_Nashorn2.jpg
“Oftheunicorn” by Special Collections, University of Houston Libraries – http://digital.lib.uh.edu/u?/p15195coll18,33. Licensed under CC0 via Commons – https://commons.wikimedia.org/wiki/File:Oftheunicorn.jpg#/media/File:Oftheunicorn.jpg
“Narwhals breach” by Glenn Williams – National Institute of Standards and Technology. Licensed under Public Domain via Commons – https://commons.wikimedia.org/wiki/File:Narwhals_breach.jpg#/media/File:Narwhals_breach.jpg
“One horned Rhino” by Krish Dulal – Own work. Licensed under CC BY-SA 3.0 via Commons – https://commons.wikimedia.org/wiki/File:One_horned_Rhino.jpg#/media/File:One_horned_Rhino.jpg
“Grey Icelandic horse” by Soffía Snæland from Njarðvík, Iceland – Hestur. Licensed under CC BY 2.0 via Wikimedia Commons – https://commons.wikimedia.org/wiki/File:Grey_Icelandic_horse.jpg#/media/File:Grey_Icelandic_horse.jpg
“Alboxmonachilmotril 039” by No machine-readable author provided. Iván Salvía assumed (based on copyright claims). – No machine-readable source provided. Own work assumed (based on copyright claims).. Licensed under Public Domain via Commons – https://commons.wikimedia.org/wiki/File:Alboxmonachilmotril_039.jpg#/media/File:Alboxmonachilmotril_039.jpg
“CanadianRusticPony” by SriMesh – Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons – https://commons.wikimedia.org/wiki/File:CanadianRusticPony.png#/media/File:CanadianRusticPony.png
“Shakudo Tsuba” by Grizan – Own work. Licensed under Public Domain via Commons – https://commons.wikimedia.org/wiki/File:Shakudo_Tsuba.jpg#/media/File:Shakudo_Tsuba.jpg
“Japanese – Tsuba with a Frog in a Lotus Pond – Walters 51177 – Back” by Anonymous (Japan) – Walters Art Museum: Home page Info about artwork. Licensed under Public Domain via Commons – https://commons.wikimedia.org/wiki/File:Japanese_-_Tsuba_with_a_Frog_in_a_Lotus_Pond_-_Walters_51177_-_Back.jpg#/media/File:Japanese_-_Tsuba_with_a_Frog_in_a_Lotus_Pond_-_Walters_51177_-_Back.jpg
“Tamagawa Masaharu – Tsuba with a Monkey Teasing an Elephant with a Stick – Walters 51281” by Tamagawa Masaharu – Walters Art Museum: Home page Info about artwork. Licensed under Public Domain via Commons – https://commons.wikimedia.org/wiki/File:Tamagawa_Masaharu_-_Tsuba_with_a_Monkey_Teasing_an_Elephant_with_a_Stick_-_Walters_51281.jpg#/media/File:Tamagawa_Masaharu_-_Tsuba_with_a_Monkey_Teasing_an_Elephant_with_a_Stick_-_Walters_51281.jpg
“Japanese – Tsuba with a Dragonfly – Walters 51254” by Anonymous (Japan) – Walters Art Museum: Home page Info about artwork. Licensed under Public Domain via Commons – https://commons.wikimedia.org/wiki/File:Japanese_-_Tsuba_with_a_Dragonfly_-_Walters_51254.jpg#/media/File:Japanese_-_Tsuba_with_a_Dragonfly_-_Walters_51254.jpg
Other notes:
You’re welcome to use Hyde & Rugg copyleft images for any non-commercial purpose, including lectures, provided that you state that they’re copyleft Hyde & Rugg.
There’s more about the theory behind this article in my latest book:
Blind Spot, by Gordon Rugg with Joseph D’Agnese
http://www.amazon.co.uk/Blind-Spot-Gordon-Rugg/dp/0062097903
Overviews of the articles on this blog:
https://hydeandrugg.wordpress.com/2015/01/12/the-knowledge-modelling-book/
https://hydeandrugg.wordpress.com/2014/09/19/150-posts-and-counting/
hyde and rugg 
By chance I came across this post. Thanks!
I had reached the same conclusion to yours about things, concepts and words, and I am in the process of writing a thesis on this.
For research purpose, is the 2016 copyleft table given in your book “Blind Spot”?
This three way distinction is well established in linguistics, so you should be able to find more about it there.
This particular table isn’t in “Blind Spot” but there’s a section in there about using this sort of permutation table systematically in research; for example, in definitions of life and quasi-life. One of my PhD students applied this type of table to medical classification of syndromes, with interesting results.
Thank you.
I am aware of this three-way distinction in linguistics (Ogden-Richards’ triangle of meaning/reference) but very few break it down into clear separate categories and analysing them individually.
Is there anyone who has already done so after them? Because my thesis is to do so.
My apologies for trying to teach you to suck eggs in my initial response; now that I’ve visited your blog, it’s clear that you’re at least as familiar with the linguistics literature as I am.
I don’t know of any further work on this within linguistics, but I haven’t followed the linguistics literature for years, so that’s not necessarily a significant absence. If you haven’t encountered anything more recent in linguistics, it might be that the linguistics community have treated this as a basic building block and as nothing more than a building block, not needing further elaboration.
You might find some different literatures highly relevant. Category theory is a very promising example, as is taxonomic theory. I’ve found that both of these give insights above and beyond what you’d get from set theory in mathematics.
For example, categories in traditional logic such as “human” can be decomposed in terms of their consituent defining characteristics, each of which may be crisp or fuzzy or rough or indeterminate etc; also, some of those characteristics may be mandatory for category membership while others may be optional. Apologies if this is also teaching you to suck eggs, but it’s highly relevant to the “concept” part of the thing/concept/word triad.
In addition, the representations that researchers use can interact with human error in various ways, with some representations reducing the risk of faulty reasoning, and others increasing it. I haven’t seen much detailed discussion about this in the error literature, though the basic concept is pretty much taken for granted there.
I hope this helps as a starting point. I’d be very happy to discuss this with you in more detail offline if you wish; from your blog articles, it looks as if we’ve been looking at much the same topics from interestingly different angles.