LOWER

What it does

LOWER(text)

LOWER returns its argument with every uppercase letter replaced by its lowercase form. Microsoft's article states both halves of that: it converts all uppercase letters in a text string to lowercase, and it "does not change characters in text that are not letters" — digits, spaces and punctuation pass through untouched.

The interesting part of the definition is what "its lowercase form" means once the text leaves ASCII. Case mapping in Unicode is not a per-character permutation:

• Some characters lowercase to a different number of characters than they started with. İ (U+0130, Latin capital I with dot above) has a canonical lowercasing of two code points, i followed by a combining dot above.
• Some lowercasings are context-dependent. Greek capital sigma Σ lowercases to final sigma ς at the end of a word and to medial sigma σ elsewhere — the same input character, two answers, chosen by what is next to it.
• Some characters have no agreed lowercase at all, or acquired one late: capital sharp s ẞ (U+1E9E) is the standard example.

A worksheet function that must return one text value per input has to take a position on all three. Microsoft's article takes none of them explicitly, so the exact mapping Excel applies is documented only by example, and the examples are ASCII.

Arguments

text — required, described in the article as the text you want to convert to lowercase. It is a text argument, so numbers and logicals convert to text first; LOWER(TRUE) is the lowercasing of the string TRUE, not a logical.

There is no locale argument. That absence is load-bearing: the Turkish dotted/dotless-i pair is the canonical case in which correct lowercasing depends on the language of the text, and LOWER gives a formula no way to say which language it is in.

Try it

Quirks

Result and edge cases

The return kind is Text. Empty text returns empty text.

Arrays, empty cells and omitted arguments follow the shared call model (chapter 03).

Errors

The article lists no error conditions, and the function has no domain to violate. The reachable errors are the shared ones: an error value arriving as the argument propagates, and a value kind with no text conversion surfaces #VALUE! under the shared coercion rules (chapter 02).

How to implement

Notes for implementers

1. Choose and record a case-mapping table version. Unicode's case mappings change between versions. An implementation that delegates to a platform's to_lowercase inherits whichever version that platform shipped, which is not a stable target.
2. Decide the expansion policy explicitly. If a mapping produces more than one character, does the function emit the expansion or leave the character unchanged? The two policies are visibly different on İ and on ẞ.
3. Final sigma needs a neighbour test. Any implementation that maps character-by-character without lookahead will get Greek text wrong at word ends.

Rust code

14 lower_fn "LOWER" (1, 1) [V "text"] |s, a, sv| { cased(s, a, k::lower) }

No implementation has been published with a Handbook test suite for this function.

Version differences

No comparison between specific Excel builds is recorded here yet. This does not establish that all builds behave alike.

Alternatives

Relationships

• UPPER is the other direction, and is not its inverse: LOWER(UPPER(x)) is not always x, and neither is UPPER(LOWER(x)), because case mapping loses information (ß and SS) and gains context (final sigma).
• PROPER capitalises the first letter of each word and lowercases the rest, and therefore contains a copy of both mappings.
• EXACT is the function to reach for when the question is whether two strings differ only by case: EXACT(a,b) is case-sensitive, a=b is not. Writing LOWER(a)=LOWER(b) as a case-insensitive comparison is a common idiom and a subtly different predicate, because it also folds any character pair that shares a lowercase form.

Named variants

No named variants are recorded yet.