# Use of the word "literally" in the definition of literally

Recently, the word literally has come to be used for emphasis, rather than to express that something definitely happened.

If you look up the definition of literally on Google, the second definition is:

used for emphasis while not being literally true.

It feels wrong to me that the word literally is in the definition. It seems like this could be interpreted as circular logic since we don't know which definition of literally the definition refers—it could refer to itself.

I am not aware of any such convention where if you use the word you are attempting to define in the second definition, then you use the first definition.

It seems to me that it would be perfectly valid to leave out the word literally from the second definition altogether, and everything would be fine, i.e.:

used for emphasis while not being necessarily true.

Is it wrong that the word literally has been used in its own definition?

• No. It could perhaps be worded "used for emphasis while not being actually literal." Aug 4, 2014 at 9:03
• I think it's pretty confusing to use a word in it's own definition. There must be a better way of wording it. Aug 4, 2014 at 9:04
• 'It seems like this could be interpreted as circular logic since we don't know which definition of literally the [second] definition refers to': I think we can make a reasonable guess. As the fine usage notes given by AHDEL, Collins and RHK Webster's say, the major problem is with the use of the word in the secondary sense, not how the second sense is defined. Aug 4, 2014 at 9:16
• I thought I'd point out that the use of literally as hyperbole isn't new. Mark Twain for example, used it. Aug 4, 2014 at 22:54
• @tchrist Not a duplicate at all. This question is regarding circular usage in a dictionary. Aug 4, 2014 at 23:42

There are two meanings of "literally".
In the definition of the second meaning, the first definition is used. So it is not circular.

"literally(1)" = [the classic definition...]

"literally(2)" = "used for emphasis while not being literally(1) true."

Of course, this relation is very implicit. But appart from that, the logic is correct.

• Why is the first definition used, though? Can you cite any source on that convention? I don't know of any other words in the English dictionary with a definition like this. Aug 4, 2014 at 9:17
• That was my interpretation (the implicit part) - and it was confirmed when I found it seems to be the only way it is not circular. I do not think it should be defined like this - I just accepted that natural language is not exact. ;) Aug 4, 2014 at 9:22
• It's poor editing on the part of the dictionary IMO, not to put the "(1)" (or similar) to make things clearer. But it's not quite as bad as the OP makes it look. If you do the google search you'll find the 2 definitions aren't actually presented as independent: the second is a bullet point under the first. And in what looks to me like the original source this is emphasised with the numbering (it's not defs 1 and 2 it's 1 and 1.1). oxforddictionaries.com/definition/english/literally
– Rupe
Aug 4, 2014 at 9:36
• @Rupe Yes - if the secondary sense is considered acceptable (and one has to be pretty prescriptivist not to), it should really be listed as a different word, its meaning being so different that there is a 'Janus word' situation. Thus cleave1 and cleave2 are listed as separate words by AHD and Collins. Divergence of meaning can trump common ancestry here (see duck1,2 at AHD). The different senses of hurricane are far less disparate (hyponymy + polysemy) and need disambiguation. Aug 4, 2014 at 9:57

Is it wrong that the word literally has been used in its own definition?

I am not aware of any such convention where if you use the word you are attempting to define in the second definition, then you use the first definition.

As is well known, dictionary definitions are inevitably circular as they are limited to a finite number of words. The 'mountain = large hill' and 'hill = small mountain' incident caused an uproar, because the circle was so small it was difficult to miss. Most circles have a few more arcs, and there are intersecting circles. Illustrated dictionaries have an advantage here.

The use of one sense of an orthographic word in the definition of another sense is by no means unacceptable and sometimes unavoidable without being ridiculous. Thus AHD uses the noun northwest in defining the senses of the adjective (or attributive noun?) northwest:

1. To, toward, of, facing, or in the northwest.

2. Originating in or coming from the northwest: a northwest wind.

The dictionary provided by Google has:

barbecue ...noun: ...

a portable grill used for the preparation of food at a barbecue

As Volker says, senses (barbecue1a) should be distinguished in such definitions.

• the mountain = large hill, large hill = mountain thing is a thing because the difference between a mountain and hill is somewhat arbitrary. There is no clear delineation between the two, there is no technical or scientific consensus if something is a hill or a mountain. The adjective definition for Northwest references itself because the adjective is just the noun in the adjective form.
– Mike
Oct 26, 2015 at 0:22
• I'm not sure how either of these comments addresses the legitimacy of the circularity of dictionary definitions. Oct 26, 2015 at 8:36