- descriptive text
- artist's statement
Longer texts like an artist's statement will tend to have more impact, even if they don't address the image directly. Having that text in mind, whatever it talks about, will surely affect the way the viewer sees the image.
Viewers will always come to an image with a large set of other material already present in their minds. We have all seen many photographs and images, we all have seen a lot of real things. We're probably been exposed to some art, somewhere along the line. We come to a photograph in a mood, with the memories of the morning drive to work, last night's dinner. All these things also affect how we will view an image, just as the accompanying text will. Text plays a role, but it can only hint, suggest, influence. It cannot control the viewer's response to the image.
Adding text to an image is an explicit acknowledgement that the image alone cannot carry the message. Perhaps the message, whatever that means, is too large and detailed (see, for instance Let Us Now Praise Famous Men) or perhaps the photograph is simply too weak. Perhaps the photograph is incapable, by its nature, of carrying important information; a documentary image might require a date and a location to serve its documentary purpose.
One can argue that the more text there is to accompany the photograph, the more the artist recognizes the inability of the photograph to carry the desired meaning. Why else have all that text? Even if the photograph exists to support the text rather than the other way around, the intent is clear: the photograph cannot carry the piece. We see shows, from time to time, of abstract or banal images accompanied by a one or more substantial texts that essentially tell us how to react to the images. Is this photography, writing, or some sort of multimedia thing? Parsing this out in any kind of detail is probably a waste of time, but it's clear that we should not view this kind of thing as purely photography.
Beware, I think, of writing too much after the fact. If your image cannot support itself without text, adding text afterwards might make something, but it won't make a photograph.