What Are Dreams? Definitions and Operational Criteria

Almost everyone dreams, and almost everyone assumes they know what a dream is — until they try to say it precisely. Is a dream only the vivid, story-like adventure you remember on waking, or does the half-formed thought you had drifting off count too? Does a dream still happen if you never recall it? And where, exactly, does a dream end and ordinary sleeping thought begin? These questions turn out to be harder than they look, and how sleep science answers them shapes everything else that can be said about dreaming. This article does one focused job: it gives the working definition and criteria that dream research actually uses — a dream as an experience during sleep, known through the report a person gives on waking — shows why dreaming is not just a feature of REM sleep, and stays honest about what is settled and what is still open. It is a definition and a set of criteria, not a theory belonging to any one thinker, and not a guide to interpreting what your dreams mean.

What counts as a dream?

In sleep and dream science, a dream is best defined by the experience and its report rather than by any single brain state. A dream is something experienced during sleep — imagery, sensations, thoughts, emotions, sometimes a whole unfolding scene — that the sleeper can describe after waking. That definition is deliberately modest. It does not say what dreams are for, where in the brain they arise, or what they mean; it simply fixes what researchers are talking about when they study dreaming. Within that frame, dreams are not all one kind of thing. At one end sit rich, immersive experiences that feel, while they last, like being somewhere and doing something; at the other end sit thin, thought-like scraps of mental activity — the bare sense on waking that you were 'just thinking about tomorrow.' Researchers often call the thinner material sleep mentation and reserve dreaming for the fuller, more immersive experiences, but the line between them is a matter of degree, not a sharp boundary. One influential attempt to pin down the minimum — the immersive spatiotemporal hallucination model proposed by the philosopher Jennifer Windt — treats dreaming as a felt sense of being present in a world during sleep. That is one useful conceptual framework among several, offered here to show how the definition is debated, not as the final word.

Dream (operational definition)

An experience occurring during sleep — imagery, sensations, thoughts or emotions, up to a whole unfolding scene — that the sleeper is able to report on waking. It is defined by the experience and its report, not by a particular brain state or by what the dream might mean.

Sleep mentation

Thinner, more thought-like mental activity during sleep, such as a bare recollection of having been mulling something over. Researchers often distinguish it from fuller, immersive dreaming, though the two shade into each other rather than dividing cleanly.

Dreaming is more than REM sleep

One of the most persistent ideas about dreams is that they belong to REM sleep — the stage of rapid eye movement first shown, in 1953, to recur regularly through the night and to line up with vivid dream recall. That discovery opened the modern, physiological study of sleep and dreaming, and for a while it made a tidy story: dreaming was what the brain did during REM. The story did not survive closer looking. From the early 1960s, laboratory studies that woke sleepers at different points in the night found that people roused from non-REM (NREM) sleep also reported dreams — sometimes less vivid, but recognisably dream experiences all the same. Later high-density EEG work reinforced the point. Dreaming, it turns out, can be gathered across the sleep cycle rather than switched on only in REM. (How the REM and NREM stages themselves work, and how they cycle across a night, is the subject of a separate Oneirica article on sleep stages and architecture; here they matter only insofar as they bear on what a dream is.)

A dream is known through the report

Here is the feature that makes dreams unusual to study: no one can observe a dream from outside. A dream is known mainly through the report a sleeper gives on waking, which makes it a report-based construct — researchers infer that a dream occurred, and what it contained, from what a person recalls and describes. That has an important consequence for anyone who suspects they 'don't dream.' Dream recall varies widely from person to person and depends heavily on how and when someone is woken; a person who rarely remembers a dream at home may produce a detailed report when woken directly from sleep in a laboratory. So not remembering a dream is not evidence that no dream occurred — absence of recall is not absence of dreaming. How often people remember their dreams, and what raises or lowers that rate, is a rich topic in its own right, and Oneirica treats it separately in its article on dream recall frequency; the point to carry forward here is only the methodological one: the dream we study is the dream that gets reported.

The search for the dream's brain signature

If dreaming is a real experience during sleep, does it leave a detectable trace in the brain? A widely discussed high-density EEG study reported that the presence of dreaming — in both REM and NREM sleep — was associated with a local drop in low-frequency electrical activity in a posterior region of the cortex, which its authors described as a posterior 'hot zone.' It was an appealing result, seeming to offer a neural signature that tracked dreaming in real time. But it is important to read such findings carefully, and honestly. A published 2020 commentary argued that the association may reflect dream recall — whether a person can report a dream — rather than dreaming itself, and that follow-up work has replicated the pattern only partially. On that view, the neural correlates of dreaming have not yet been identified so much as proposed. This is exactly the kind of finding that should be held at arm's length: genuinely interesting, actively researched, and not settled. The honest summary is that dreaming is associated with distinctive brain activity, that a posterior 'hot zone' is one candidate, and that no neural signature of dreaming can yet be treated as established.

Why do we dream? Theories, not consensus

Defining what a dream is does not settle why we have them, and on that larger question dream science has no agreed answer. What it has instead is a set of competing models, each illuminating and none established as the consensus. The protoconsciousness hypothesis, associated with the sleep researcher J. Allan Hobson, links REM sleep and dreaming to a primitive, built-in form of consciousness that the waking brain builds upon. Windt's immersive spatiotemporal hallucination model comes at dreaming from philosophy, asking what minimally must be present for an experience to count as a dream at all. The neurocognitive theory developed by G. William Domhoff treats dreaming as the output of cognitive systems and emphasises that dream content, studied across many people, tends to echo the concerns and activities of waking life — a broad pattern often called continuity. That continuity is a group-level tendency, not a code: it does not license tracing a particular dream back to a particular waking cause, and neither a universal nor a personalised dictionary of dream symbols has been empirically established as a way to decode an individual dream. Each of these accounts explains part of the picture; whether any of them explains why we dream remains an open scientific question.

Common misconceptions

• That dreams happen only in REM sleep. Dream reports have been recovered from NREM sleep since the 1960s; REM is strongly associated with vivid dreaming, but it is not the sole home of dreams.
• That not remembering a dream means you did not dream. Recall varies widely between people and with how you are woken; low recall is not the same as an absence of dreaming.
• That science has pinpointed where dreams come from in the brain. Distinctive brain activity is associated with dreaming, and a posterior 'hot zone' is one proposal, but no neural signature of dreaming is yet established.
• That there is a fixed, universal dictionary of dream symbols. No such symbol key has been empirically validated; broad continuity with waking life is a group-level pattern, not a code for reading a single dream.
• That one theory has settled why we dream. Several serious models compete, and the function of dreaming, if it has one, remains genuinely unresolved.

How to think about your own dreams

The most useful thing this definition offers is a kind of calibrated honesty. A dream is a real experience — something genuinely happened in your mind during sleep, and it is worth taking seriously — but the deeper questions of why it happened, what it means, and how it maps onto your brain are, scientifically, still open. Holding both halves at once keeps you clear of the two common traps: dismissing dreams as meaningless noise, and over-reading them as coded messages waiting to be decrypted. If you want to go further on Oneirica, related topics pick up the threads left loose here: the sleep stages and architecture that dreaming rides on, why dream recall varies so much from person to person, the special case of becoming aware you are dreaming in lucid dreaming, and what current research can and cannot say about why we dream at all. A dream, in the end, is easier to have than to define — but a clear, modest definition is what makes everything else about dreaming possible to study.