There is no research support for a fixed meaning per dream image. What research does support is that dreams track waking preoccupation. That makes the lookup table the wrong tool rather than the subject a waste of time.
A dream dictionary is an alphabetical list of images with a meaning attached to each: water means emotion, teeth mean anxiety about appearance, a snake means transformation or betrayal depending which book you opened. The format promises that a dream can be decoded the way a foreign sentence can, one word at a time.
The problem is not that the meanings are strange. It is that they are fixed. A dictionary entry has to give the same answer to everybody, and there is no evidence that a dream image means the same thing across two people, or across two nights for the same person. The evidence points the other way: what an image is doing in a dream depends heavily on who is dreaming it and what their week has been like.
It is worth being fair about why the format persists. It answers the question the reader actually asked, immediately and at no cost, in a form that fits on a phone screen at six in the morning. Any honest alternative has to accept that it is offering something slower and less satisfying, a set of considerations rather than an answer, and pretending otherwise would be its own kind of dishonesty.
This site publishes thirty pages about specific dream images, so the position needs stating plainly. Those pages are not entries in a lookup table and are not written as one. A page on teeth falling out sets out what the image has meant to different traditions, what the psychological schools said about it, and what content research has found about how often it actually turns up and in what company. The reader supplies the part a dictionary pretends to supply, which is what the image was doing in their particular dream.
The distinction is between a page that tells you what your dream means and a page that tells you what people have thought this image means, and then hands the interpretation back to you. The first cannot honestly be done. The second is ordinary cultural history, and it is worth reading.
If not a lookup, then what. The approach with the most support behind it is unglamorous and takes about four minutes.
Write the dream down before you have explained it to yourself, in the present tense, including the parts that make no sense. Then, separately, write what the last two or three days have actually been like: what you were worried about, what you were putting off, who you had a difficult conversation with. Then look for the join. The continuity hypothesis predicts that one will usually be there, and in practice one usually is.
Two rules make this work better. Record the feeling as well as the image, because the emotional tone of a dream is more consistent and more informative than its contents. And record a series rather than a single night: patterns across a fortnight say considerably more than any one dream, a point Jung made on theoretical grounds and content researchers later reached by entirely different means.
What you are producing is not a translation. It is a record of what your attention has been on, written by a part of your mind that does not edit for presentability. Keeping a dream journal sets out the practice in detail.
The best-supported finding in dream research is the continuity hypothesis: dream content largely continues waking concerns. It comes out of the content analysis tradition begun by Calvin Hall in the 1940s and 1950s, and it has held up across decades of coded dream reports. People dream about the people they know, the places they spend time in, and whatever is currently occupying them.
The continuity is selective rather than literal, which is a detail worth having. Dreams substantially under-represent activities that fill a great deal of waking time, including reading, writing and screens, and over-represent interpersonal situations and emotionally charged ones. What continues is preoccupation rather than activity.
The systematic evidence comes from the coding system Calvin Hall and Robert Van de Castle published in 1966, which established norms from a thousand dream reports and made dream content countable. It found strong population regularities: most dreams contain at least one other character, aggression appears more often than friendliness, misfortune more often than good fortune. G. William Domhoff later extended and computerised the method, and argued from the same data that dreams are dramatised expressions of a person's conceptions and concerns, with no evidence of symbolic disguise anywhere in the record.
The most substantial theory about why dreams take the shape they do is Antti Revonsuo's threat simulation hypothesis, proposed in 2000, which reads the over-representation of threatening situations as rehearsal: dreaming evolved, on this account, as a way of practising responses to danger. The underlying observation that threat is over-represented is well supported. The evolutionary explanation for it remains contested. Note what the theory is not, though: it is an account of what dreaming in general might be for, not a key to individual images, and it would not license a dictionary even if it were settled.
What none of this work found was a fixed meaning per image. The norms describe how often something appears and alongside what. They do not assign a significance to a snake. Nothing in the research literature supports the proposition that a given image carries a stable meaning independent of the dreamer, and that is the point where dream dictionaries and dream research part company completely.
The traditions disagree with each other, which is the first evidence against a fixed table. Artemidorus, writing in the second century CE, gives one reading for a dreamer who is a craftsman and a different one for a dreamer who is wealthy. He also separates dreams that show a thing plainly from dreams that show it obliquely, and treats the two as requiring different handling, which is already a more careful apparatus than any modern dictionary offers.
Medieval European dream books give flat lists. Islamic dream interpretation in the ta'bir tradition treats the dreamer's circumstances and character as central to any reading. Classical Chinese dream literature reads images against the dreamer's constitution and the season.
Where two traditions do agree about an image, the agreement is usually cultural rather than universal. A snake reads differently in a place where snakes are a daily hazard than in one where they are an animal you pay to look at.
The psychological schools disagree just as sharply. Freud held that dream images disguise their meaning; Jung held that they do not. Freud and Jung on dreams sets out how far apart the two of them actually were, and it matters here, because dictionary sites routinely blend them into a single voice neither man would have recognised.
The genre is old, and its founding text argues against it. Artemidorus of Daldis compiled the Oneirocritica in the second century CE, and it is the ancestor everybody cites. What it actually says is that the same dream means different things to different dreamers, and that an interpreter must know the dreamer's occupation, status, habits and health before saying anything at all. He is explicit and repetitive on the point.
The alphabetical fixed-meaning format arrives later. The Somniale Danielis, a dream book attributed to the prophet Daniel, circulated widely across medieval Europe in exactly the shape modern dictionaries use: an alphabetical list of images with one meaning each and no dreamer required. That is the real ancestor of the format.
The commercial phase is nineteenth century. Cheap dream books sold in enormous numbers, and in the United States many were explicitly numbered, so that a reader could convert a dream image into a figure to bet on in the numbers games. The dream dictionary and the lottery slip share a printing history.
Freud's Die Traumdeutung appeared in 1899 and gave the whole subject intellectual respectability, which the dictionary trade absorbed without adopting any of his method. Most modern dream dictionaries are the medieval format wearing psychoanalytic vocabulary.
The twentieth century industrialised that format into mass-market paperbacks, and the internet industrialised it again. A dream dictionary is close to ideal material for a search engine: thousands of distinct queries, each answerable in two hundred words, none of them checkable by anybody. That is a large part of why so many of these sites exist, and it is a fact about search economics rather than a fact about dreams.
The first misreading is taking this argument as debunking. It is not. Dreams are real events with measurable structure, they track what people care about, and thinking about them is a reasonable thing to do. What has no support is the specific claim that an image carries a fixed meaning, and rejecting that claim leaves almost everything interesting still standing.
The second is the opposite error, which the genre actively encourages: treating a dictionary entry that happens to resonate as confirmation the system works. Entries are written broadly enough to resonate with most people most of the time, and recognition is not evidence. The real test would be whether an entry tells a dreamer something about themselves they did not already know, and nobody in the trade publishes that test.
The third is assuming the great interpreters would have endorsed the lookup. They did not. Artemidorus insisted on knowing the dreamer before interpreting anything. Jung wrote that no dream symbol can be separated from the person who dreamt it, and was scathing about mechanical translation. The two authorities most often invoked on dream-dictionary sites both explicitly rejected the method those sites are using, which is worth knowing before paying anybody for a reading.
Look up what you actually dreamed.