We spend roughly a third of our lives asleep, and about a quarter of that time dreaming. Most of those stories fade within minutes of waking, which is why dreaming stayed at the edge of serious science for so long. A team in Italy decided to count them anyway: more than 3 700 dream reports from almost 290 people, run through language algorithms. What came out is a picture of dreaming that is anything but random — and of dreamers who differ from one another enough that it is worth asking what kind of dreamer you are.
Sleep takes up about one third of a human life, and roughly a quarter of that time is spent dreaming. For most people the content evaporates within minutes of opening their eyes, which is part of the reason dreams sat at the margins of laboratory science for decades: the material is real, but it disappears before anyone can examine it.
Ancient cultures read dreams as messages from the gods. The data is less flattering. Dream reports turn out to be dominated by negative emotions far more often than by pleasant ones.
Researchers at the IMT School for Advanced Studies Lucca in Italy took the laborious route. Between 2020 and 2024 they collected more than 3 700 dream accounts from almost 290 volunteers aged 18 to 70, then processed the texts with natural language processing. Instead of interpreting single dreams, they looked for patterns across thousands of them at once.
Valentina Elce, who led the work, describes the sleeping brain as a system cut off from its usual supply of input: "sleep is a state where the brain is more disconnected from the environment compared to wakefulness. So when we're sleeping, our brain basically is free to explore distant concepts and connections with minimal interference from the outside world. It's processing all the information that we've acquired during the day to allow us to learn new things, to consolidate our memories."
The analysis showed that dreams lean heavily on familiar settings, but hardly ever reproduce them. Places are recombined, timelines slip, people show up where they never were. The raw material comes from the day. The arrangement does not.
That distinction matters, because it separates two very different ideas of what sleep does with memory: a recording, or a rewrite.
Individual differences turned out to be readable in the text of the reports themselves. Among the patterns the team found:
In other words, a dreamer's profile is not a mystery to be decoded symbol by symbol. It lines up with traits that are already visible while the person is awake.
The second result is methodological. Software analysed the dream narratives about as reliably as trained human scorers, which removes the bottleneck that has limited this field from the start: until now somebody had to read every single report by hand.
The findings were published in the journal Communications Psychology. Elce sums them up this way: "Our findings show that dreams are not just a reflection of past experiences, but a dynamic process shaped by who we are and what we live through. By combining large-scale data with computational methods, we were able to uncover patterns in dream content that were previously difficult to detect."
Student years pile up exactly the kind of input dreams feed on: new places, exam pressure, deadlines and sleep that shifts from week to week. If dreaming is a processing stage rather than background noise, then a string of broken nights is not only a question of feeling tired the next morning.
There is a practical footnote too. Dream recall improves with the habit of writing things down shortly after waking, which is the same reason the Lucca team was able to gather thousands of reports in the first place.
Dreams are built out of everyday life but rarely repeat it. More than 3 700 reports from almost 290 people showed that fragmented dreams go together with a wandering mind during the day, and that emotionally charged dreams are more common among those who consider dreams meaningful. The other conclusion is technical: language software read the narratives roughly as well as human analysts, which is what makes dream research on this scale possible at all.
5 questions · one minute · nothing is saved
EN
CS
DE
ES
FR
HU
IT
PL
PT
РУ
SK
TR
УК
AR
中文
