Defining words: harder than it sounds

Publication date
Thursday 29 April 2021

There are some common problems with defining words, and some tools and strategies to solve them.

In government writing we often need to define words. This can be because the word is a proper noun used in government, or because the word is uncommon, confusing, or jargon.

If you think you have a good sense of what good and bad definitions are, some principles from linguistics can help you ground that.

Avoid obscure or circular definitions

Here is a definition of ‘fire’:

The active principle of burning or combustion, manifested by the evolution of light and heat.

Linguists call this kind of definition ‘obscure’. The definition itself uses complicated or uncommon words.

This definition may be fine for an academic who wants to record the exact meaning of ‘fire’.

A person who does not know what ‘fire’ means probably can’t understand this definition.

Another common problem is that definitions can be ‘circular’.

Important: of much significance or consequence.
Significant: important; of consequence.

These definitions are circular. Each definition refers to the other word.

If you know one of these words, circular definitions can help. If not, they are useless.

Define words for many types of users

Consider your audience when writing definitions. Consider what challenges they may face.

The users of government content often include people with low English language ability. Government content is often translated into other languages. For these reasons, we should avoid obscure or circular definitions. Some dictionaries still use circular or obscure definitions, and you should be aware of this when using dictionaries.

You can define words using simpler words. The Style Manual shows us an example of this approach in the definition of immunisation on the Easy Read page.

Use the simplest words

This approach raises 2 important questions:

  • How do we know if one word is simpler than another?
  • If we keep defining words using simpler words, won’t we hit undefinable basic words?

Linguists propose using semantic primes to solve this.

Semantic primes

Semantic primes are like prime numbers. You can’t divide a prime word into smaller words. There are 65 prime words, including ‘people’, ‘think’, ‘big’, ‘want’ and ‘place’. They also have direct translations in every known language.

Linguists test how difficult it is to define words using semantic primes. The more words and phrases needed to define a word, the more complicated it is.

A long definition using simple words may help more than a short, obscure or circular definition. For example, a less obscure way to define ‘fire’ is to say:

Something that happens when things get very hot. There is light which is often red or yellow, and the hot thing is destroyed or changed.

A less circular definition of ‘significant’ is:

An idea, amount of something, or event can be significant. If something is significant, people notice this thing more than other things. People notice a significant thing because it can make a big change to something people care about.

Linguists have used semantic primes to create a list of the 360 simplest English words. Using this list is another good way to choose simple words. There’s also an automatic minimal English checker.

Common words are not always simple

Another tool for choosing simple words is a simple English dictionary. These usually list the most used words in English. Common words are not always simple. For example, ‘awkward’ and ‘put’ are very hard to translate or define, because context is very important to their meaning. Simple English dictionaries can be a good starting point, but do not assume a word is easy to understand for all users because it is common.

Michael Turvey is part of the DTA Digital Cadetship Program, studying digital humanities and linguistics.

Help us improve Style Manual

Did you find this page useful?
Do you have any other feedback?
Is your feedback about:
Select the answer that best describes your feedback:
Do you work for government?
Are you interested in taking part in Style Manual user research?
Please tell us a bit more about yourself.
Do you work for government?