History of Dreaming

In the late 1920s, the psychiatrist Hans Berger introduced a new technique to detect the cortical activity from the scalp of human being. In 1929, the physician showed the first registration of the cortical electric activity obtained by means of electroencephalography machine, from which it was possible to get an electroencephalogram (EEG) on printed paper. In subsequent years, such technique was also adopted for sleep recordings (Loomis et al., 1937), revealing that the electric activity was not homogeneous, but instead, it changed across each night. For several decades, ocular movements observed during sleep were considered casual and insignificant. However, Eugene Aserinsky, a physiology student at the University of Chicago under the supervision of Nathaniel Kleitman, revolutionized sleep scientific research by measuring cerebral activity and ocular movements simultaneously through electrooculography (EOG). They discovered specific intervals with rapid and recurrent eye movement and burst of alpha activity, comparable to those during wakefulness (Aserinsky e Kleitman, 1953). They named this particular kind of sleep as Rapid Eye Movement (REM), which for its features (i.e., desynchronized EEG activity, muscular atonia and rapid eye movements) is also known as "paradoxical sleep". On the other hand, the Non-Rapid Eye Movement (NREM) sleep consists of 1 and 2 lighter sleep stages and 3 and 4 deeper sleep stages and exhibits slow and synchronized EEG activity. Essentially, the year 1953 coincides with the birth of psychophysiology of sleep. Until then, psychoanalysis had the primacy in the study of dreams. Freudian theories about their interpretation considered the oneiric activity as a doorway to access the unconscious functions of mind in the neurosis care (Freud, 1899). From 1953, the enthusiasm for the discovery of REM sleep considerably influenced the following research about dreaming: it was developed the idea that the specific physiology of REM sleep would generate oneiric activity, since it was being observed the highest probability to find the dream recall at awake from this stage. IndeedAserinsky e Kleitman (1953b) , in their first studies, found dream reports with hallucinatory, complex and bizarre content in 74% of awakes from REM sleep. On the contrary, the same results were found only in 9% of awakenings from NREM sleep (Eiser, 2005), in which mental activity with poor mental images was more frequent, without structured plot or, in some cases, the impossibility to recall the dream (Horne, 1993). For several years thereafter, almost the totality of studies was set up on the biunivocal correspondence REM=dreaming (Hobson & McCarley, 1977; Hobson, 1988; Maquet et al., 1996; Braun et al., 1997; Nofzinger et al., 1997; Hobson et al., 2000). In fact, because of the specific physiological differences between sleep stages, many researchers considered REM sleep as the neural correlate of dreaming. Researchers asked the subjects if they were dreaming something before awakening, influencing in this way the answer. Indeed, any individual could give a different subjective interpretation to the word "dream". Bearing in mind this consideration, in 1962 the psychologist David Foulkes realized that it was possible to obtain oneiric report also after awakenings from NREM sleep, just modifying the question and using more liberal criteria. More generally, Foulkes asked the participants if “anything passing through your mind” (Foulkes, 1962; Horne, 1993). After several in-depth analyses in this direction, two fundamental studies questioned correspondence REM=dreaming. Antrobus (1983) and Foulkes & Schmidt (1983) reported that sleep mentation occurred for the whole night, without any qualitative or quantitative difference between REM and NREM sleep. Indeed, dreams occurred in any sleep stage, during the slow wave activity and the sleep onset or the relaxed wakefulness (e.g., hypnagogic and hypnopompic hallucinations; Cavallero, 2000) and, sometimes, they could be longer in NREM than in REM sleep (Foulkes & Schmidt, 1983; Cavallero et al., 1990). Currently, the correspondence REM sleep=dreaming is still present, but it is transformed in a long debate about different mechanisms of dream's generation (Nielsen, 2000). According to many neuropsychological, neuroimaging and EEG studies, it can be now stated that mental activity occurs for the entire period of sleep (Foulkes, 1962; 1967), although with some differences which concern dream-like features (many characters, vivid images, structured and bizarre plot), on one side, (Foulkes & Rechtschaffen, 1964; Foulkes, 1967; Foulkes & Schmidt, 1983; Antrobus, 1983; Casagrande, Violani & Bertini, 1996; Stickgold, Pace-Scott & Hobson, 1994; Waterman, Elton & Kenemans, 1993) and thought-like features (vague and few images; Foulkes, 1967), on the other. In this contest, some researchers hypothesized that dreams produced from REM sleep would be more dream-like and that dreams generated from NREM sleep would be more thought-like and proposed the so-called Two-Generation Model (2-gen model; Hobson, Pace-Scott & Stickgold, 2000; Nielsen, 2000). Nevertheless, this hypothesis was not ever being demonstrated (Antrobus, 1983; Bartolacci, Scarpelli & De Gennaro, 2017). Indeed, many vivid and hallucinatory dreams also occur in NREM sleep, as pavor nocturnus from stages 3 and 4 (Nielsen, 2000; Fisher et al., 1970a; 1970b; 1973; Kahan, Fisher & Edwards, 1991) or hypnagogic and hypnopompic hallucinations in transient states sleep-wakefulness (Nielsen, 2017). For these reasons, assuming the homogeneity of mental activity during the sleep, the One-Generation Model has been proposed (1-gen model; Solms, 2000; Foulkes, 1962; 1967), that will be discussed more in detail below.