God Didn't Say That

Bible Translations and Mistranslations

Allah and the Type-Token Distinction

As has been widely reported recently, Malaysians are grappling with how to say “god.”

The issue is that the word most Malay speakers use, allah, refers not only to “god” in general but in particular to what we call “Allah” in English, that is, the Muslim god of that name. (For the blow-by-blow, start here for background, then see here for a recent court ruling reversing an original decision and allowing anyone to use the word allah. And see here for a report of massive violence against churches for using the word allah in Christian contexts. See also here for the court’s suspension of its reversal of its original ban.)

The problem essentially stems from type-token conflation, and not for the first time.

Essentially, there is a category (“type”) that in English we usually spell with the lower-case-g word “god,” and for which we can equally use the word “deity.”

In addition to the category, there is what the category contains (“tokens”). One example of such a token is what we usually write with the upper-case-G word “God” in English.

The distinction is usually pretty clear. For example, the type “Greek god” contains the tokens “Zeus” and “Athena,” among others.

But monotheism involves a type that has exactly one token, so it’s pretty easy to use the type and token interchangeably. We do this in English when we talk about “praying to God.” Translations even stumble sometimes, prefering the upper-case-G “God” where the lower-case version makes more sense, as in “our God” for “our god.”

We see the same interchangeability in the Hebrew of the Old Testament. The word elohim is used both for the God of the Hebrews and for any other god. And the Greek theos, represents variously “God” or “god” (or “goddess”).

Similarly, the word allah (before Islam merely one Arabian god of many) is now the Muslim name for “God” as well as the type “god.”

If nothing else, I think we see here that a solid theoretical framework, while essential for translation, is only part of the story. (I don’t think the end to the crisis in Malaysia is a campaign explaining the nature of type-token conflation.)

And I wonder if the authors of the Bible similarly debated the words elohim and theos.


January 11, 2010 - Posted by | general linguistics | , ,


  1. Nice one Joel
    though I think when Christians however nominal use an uppercase G in “our God” they may bew stumbling in terms of type-token confusion, but actually the sematic referent is obvious both to the Christian and to the non Christian.
    This also has a bearing on what’s happening in Malaysia as muslims will only allow Allah” to refer to (as they would say) “our God” and not “your god”.
    Don’t think I’m disagreeing, perhaps just saying the same thing in a different way
    cracking photo btw

    Comment by Tim Goodbody | January 11, 2010

    • I agree that what I’ve called type-token confusion doesn’t usually result in any actual confusion.

      Comment by Joel H. | January 12, 2010

  2. ISTM that a great deal of capitalization in the translations is prejudicial, lending inappropriate suggestions and should be eliminated. Your excellent example of “our God” is one. We see it also in references to “Son” and “Spirit.” This allows the unscrupulous translator to infuse Trinity dogma onto texts that would otherwise not make that suggestion at all. Hence, the use is political, rather than a faithful rendering of the underlying language.

    Comment by WoundedEgo | January 13, 2010

Sorry, the comment form is closed at this time.