Ask Marilyn ® by Marilyn vos Savant is a column in Parade Magazine, published by PARADE, 711 Third Avenue, New York, NY 10017, USA. According to Parade, Marilyn vos Savant is listed in the "Guinness Book of World Records Hall of Fame" for "Highest IQ."
In her Parade Magazine column of August 28, 2005, Marilyn is asked why mountains and sunsets appear smaller in a photograph than they do in real life, but other subjects (such as people) do not. Marilyn replied that cameras do not have peripheral vision, as people do, and suggested that the use of a wide angle lense would include more of the scene.
The questioner was concerned that the sunset or mountain range, that did appear in the picture, appeared too small--not that not enough of the scene was captured in the photograph. Using a wide-angle lens will only make it look smaller. To take the sun in the sunset as one example, it is the same size in the sky as the moon; each looks pitifully small when taken with a normal lens. To get the sun to look as large as it did in the original scene, you'd need a telephoto lens, not a wide-angle one; the latter would just make the problem worse.
As to the question of why, it's not a matter of peripheral vision, as that again addresses the question the reader did not ask, about the included objects, rather than the sizes of the objects included. The actual question the reader asked has two components to its answer: physical (optical) and psychological.
In order for the photographic image of the object to subtend the same angle as the original subject, the ratio of the frame size of the uncropped print to the viewing distance from eye to print must match the ratio of the corresponding frame size measurement within the camera to the lens's focal length. Say the lens's focal length is the 50 mm that's normal for a 35 mm camera, and the frame is the typical 24 mm x 36 mm. Larger or smaller format cameras have similar ratios for normal lenses. A typical print is, say, 6 inches long; it would have to be viewed from only 8.33 inches from the eye in order for the angles subtended by objects in the field of view to match their original values, but that is too close to the eye for most people to focus. A better solution would be to view the print from, say 16 inches, but make the print 11.52 inches long (and 7.68 inches wide).
That takes care of the optical reason for the discrepancy, but there's still a psychological reason. Even if you made the print as described in the preceding paragraph, 11.52 inches by 7.68 inches and viewed it from 16 inches away, the sun in that sunset would not look as big as the sun in the original sky. That's a constancy phenomenon in which the mind knows that the object is only inches away, while the original sun was much farther away (actually only considering it 100 or 200 feet away, not its actual distance, as it's usually described as looking the size of a basketball or some similar measurement). The brain then sizes it down from the original impression at the live scene, even though an equivalent angle is subtended by the image as by the real sun. I similar phenomenon occurs when viewing the stars in a planetarium: the subtended angles are the same at the eye, and indeed have to fit in 180 degrees across the sky and 360 degrees around, but the real constellations look much much bigger.
As to why the person in the picture looks more normal, I'm sure that has more to do with attention effects and familiarity with portraits. Besides, in the original scene they were closer to begin with so the psychological portion of the discrepancy is less.