Disjunctive cognition and interobjects: 
What psychoanalytic dream study can tell us about the brain

Mark J. Blechner

[Paper presented on June 15, 2002, at the Annual Meeting of the Rapaport-Klein Study Group]


I would like to present to you some of my ideas about how clinical work with dreams can provide data on the structure and operation of the brain. Freudís book The Interpretation of Dreams was not just about dreams. It was an attempt to develop a model of how the mind works. Freud used dream phenomena as evidence for the functions and structure of the mind. He looked at the forgetting of dreams and developed the concept of repression. He looked at the imagery of dreams and developed the concept of regression. Now, one hundred years later, armed with more clinical experience and greater knowledge of the brain, we need to extend Freudís studies, with closer examination of the phenomenology of dreams in relation to brain function.

I have been working on this project for several years. Many of my first findings are recorded in my book The Dream Frontier (Blechner, 2001). The more I work on this problem, the more I realize that I cannot do it alone. I need the collaboration of other analysts who are also interested in the theory of how the mind works, in the tradition of David Rapaport and George Klein. So you can see why I was so delighted by the invitation to speak to this group. Each one of us, as clinicians, collects data unlike any other scientist. We hear more dreams every day and in greater detail than anyone else in our society. If we pool our data and systematize our observations, we may discover things about the operation of the brain that are quite different from what laboratory neuroscientists are discovering and just as important.

Freudís model of regression in dreaming implied a one-directional flow of information while we are awake Ė from sensory contact with the world through the perceptual system and into the memory system. (You may remember Freudís flowchart that used arrows, to show the information moving from the system Sens. ŗ to the system Pcpt. ŗ and then to the system Mnem.) Freud felt that in dreaming, this flow is reversed. Input from the sensory system is blocked, and the flow of information between the perceptual and memory systems is reversed. The perceptual system receives input from images stored in memory.

Freudís proposal was brilliant in its time, but it needs to be revised today in light of new information about the mind/brain. We know that the transformation of raw sensory data into percepts is more complex than Freud thought. It occurs in many stages, and information of differing complexity is processed in different parts of the brain. The thinking today, among many scientists, is that mental functioning is best described by a model of parallel distributed processing (PDP). This means that information does not flow only in one pathway; instead there are multiple pathways for information that can shift independently (McClelland & Rumelhart, 1986).

For example, in vision, the information enters the eye, but it is divided along several pathways that lead to different parts of the brain. One part of the brain may analyze certain kinds of features, while another part of the brain analyzes the image for a different kind of information. In addition, there are many recurrent feedback loops along the way that feed the partially processed information back into the system.

So Freudís proposal that, in dreaming, the flow of information is reversed, probably needs to be revised. How can you simply reverse a process with multiple lines of information-flow and with recurrent feedback loops? But we can modify Freudís idea and propose that in dreaming, the flow of information, if not simply reversed, is changed. But how is it changed? That is a great research project still to be done, although we already have some fascinating early results, from brain-imaging studies (e.g., Braun et al., 1998).

In my view, it is not only up to neuroscientists to explore this issue. We practicing psychoanalysts must study dreams closely, for the varieties of experience that occur in dreams and the implications of these dream phenomena for understanding the organization of the brain. If we listen to dreams with neurobiological questions in mind, we may be able to outline some neurological processes based on the phenomenology of dreaming. We can identify and catalogue the kinds of cognitions that are common in dreamlife and relatively absent from waking life. Then, we can take these observations and formulate proposals about the kinds of processing of information and emotion that differentiate dreams from waking mental life. Such proposals could then help guide neurophysiologists about where to look, about which mental functions may be functionally and neurologically separable from one another, and how, in sleep, these mental functions are reorganized.

I donít have time to give you all my thoughts about how we already can do this. But I want to outline two kinds of dream phenomena, which I call disjunctive cognitions and interobjects. Both of these phenomena point the way to important aspects of brain organization.

Disjunctive cognitions

First, about disjunctive cognitions: Consider the following bizarre dream formation. A dreamer says, "I knew she was my mother, although she didnít look like her." Any clinician who hears dreams regularly knows that such statements are not at all uncommon.

It surprises me that this is not usually surprising to people. Many people, when reporting a bizarre experience in a dream, will prepare the listener by saying, "It was the strangest thingÖ" or "I donít really understand how this could happen, and yetÖ" But when people see someone in a dream whose identity doesnít match their appearance, they usually donít use a qualifying preface to describe the experience Ė at least my patients donít. They take it for granted that I will know what they mean. It is what I call "a commonplace bizarreness of dreamlife."

The dreamer recognizes a characterís identity, even though the personís appearance does not match the identity. There is a disjunction between appearance and identity. In waking life, most sane people would assume that they mis-saw or mis-identified the person, and correct for it; but not necessarily in dreams.

This is just one example of what I call "disjunctive cognitions." Two aspects of cognition do not match each other; the dreamer is aware of the disjunction, yet that does not prevent it from remaining [Footnote 1: This may be an example of what Fodor (1983) has called "The Modularity of Mind." Fodor does not consider dreams, but the modularity of mind during dreaming and dream remembering are topics worthy of study].

The global hypothesis, and I wish to emphasize, this is just a hypothesis, which I suggest we explore is as follows: Wherever disjunctive cognitions occur, the two aspects of cognition that are disjunctive are handled in different brain areas whose mutual integration is suppressed or shifted during sleep. Thus, the specifics of bizarre dream experiences may clarify what are the different stages of perceptual processing. As these are clarified, we can update Freudís regression model, and spell out how different modules of perceptual processing can interact and provide input to create the dream.

We also will want to pay attention to which disjunctive cognitions do not occur. For example, although it is common to dream, "I knew it was my mother even though it did not look like her," the converse dream is not common: "I knew it was not my mother, even though it looked like her." In fact, in all of my records of my patientsí dreams and my own over 25 years, I donít have a single dream like that. If anyone has a verbatim record of a dream like that, I would be very interested to hear of it. If such dreams do not occur or are very rare, that will also be important for our theory.

Similarly, it is quite common to dream that as an adult, one goes back to a time and place of oneís childhood. In this case, the perceived age of the dreamer is disjunctive with the setting of the dream. I was talking about this to David Olds a while ago, and he suggested that in his experience, the opposite is quite rare Ė that is, of someone dreaming of himself as a child, but where the time and place are the present. However, it is common to dream of others at an earlier age, appearing in the present [Footnote 2: See for example: Aharon Appelfeld, "The Kafka Connection," New Yorker, July 23, 2001, p. 41: "A few days after my return [from Prague], I dreamed about my parents. They had not aged since we were together sixty-three years ago in Prague, and their faces expressed amazement that I had grown older. We were briefly united in mutual astonishment, and I knew that I had something important to tell them. But, as in every profound dream, I could not get the words out." The others are their age in the past, while the dreamer is his current age. This is especially common in dreams of people who have lost close relatives. In The Dream Frontier (p. 287), I describe a woman who dreamt: "In the dream, I am my current age (65) and my mother is 32. She is ignoring me." Although she knew that this combination of ages was impossible in the waking world, it felt compelling to her. She explained: "My mother was 32 on Kristallnacht in Austria. I was only three years old. I cannot remember anything from that time. But my relatives tell me it was horrible. My uncle came to our apartment with his face all bloodied. My mother left the house to search for my father. He had been sent to a concentration camp, but she was able to bring him back. I have no memory of the time, but I think that whatever happened then has stayed with me forever. In some sense, my mother, as I react to her and to her stand-ins in my life, is always that 32-year-old woman, no matter what my actual age."].

Usually, clinicians think of disjunctive cognitions in terms of their psychodynamic meaning. Many of you may know Fosshage and Loewís (1987) book on dream interpretation, in which they presented one patientís dreams to six clinicians from different schools of thought for interpretation. The first dream included the following disjunctive cognition: "My mother was in there going through my purse. She didnít look like my mother."

Most of the clinicians, regardless of their school of thought, agreed that the person who does not look like who she is represents not the person herself but the internal object, the internalized mother. The focus among these analysts was not the division of the brain function, but the division between actual people in the outer world and internalized "people," the internal object representations, that we all carry in us.

But here we face an intriguing possibility. The fact that we can divide our perceptions between internal and external object representations must have a mechanism in the brain. What is the brain basis for internal object-representations? Is the neural anatomy that allows internal object-representations related to the separate brain areas involved in feature perception and identity perception? For anyone interested in a neurally-grounded psychoanalytic science, this seems to be an essential question, even though we may not yet have definitive answers, given our current neurological knowledge.

We may ask, isnít it odd to feel so certain about the identity of the person, in the face of contradictory physical evidence? What does this division tell us? It suggests that the processes of seeing the physical attributes of a person are not identical, perhaps not even isomorphic, with the recognition of the identity of that person. But is there any other evidence that those processes are separate in our minds?

Indeed, there is. Let us first look to the field of neuropsychology. We find that some people who have suffered strokes or other brain damage have a syndrome known as prosopagnosia. A prosopagnosic man may look at his wife of 50 years, see all of her features clearly, and yet not recognize who she is. In such people, the process of seeing is intact, but the process of facial identity recognition is not (Bodamer, 1947).

We also have the phenomenon of Capgrasí syndrome, in which a person may feel that a close relative is actually an impostor. The features are recognizable, but the personís identity is not. And there is also Frťgoliís syndrome, in which a person may mistakenly identify strangers as people that he actually knows.

We also can find relevant data from research on visual perception in animals and humans. Scientists are identifying the parts of the brain that are responsible for different aspects of face recognition (see Mesulam, 1998). Gorno Tempini et al. (1997) found that in humans, identifying unfamiliar faces activates unimodal visual association areas in the fusiform region while the recognition of familiar faces also activates an area in the lateral midtemporal cortex. Also, a similar division of function was found by Perrett et al. (1982) in macaque monkeys. From such findings, we have come to recognize that the process of facial recognition is indeed very complex and may be achieved by a part of the brain that is different from the brain areas involved in general visual analysis [Footnote 3: See Farah, 1995, for a thorough discussion of whether facial recognition is merely an aspect of very complex visual analysis or is indeed "special," and compare with Tranel et al., 1988].

We have thus a case of several kinds of data converging on a single phenomenon; we have data from human neuropathology, from experimental brain research with humans and animals, and from psychoanalytic observation of dreams all showing that the processes of feature perception and identity recognition are separated. In normal waking consciousness, the two work in tandem. That may be one of the reasons it is hard in analysis for people to become aware of their "internal objects." But perhaps the person-recognizing module of the brain can see mother even when the physical features show "not mother." In dreams, psychosis, neuropathological syndromes, and very intense transference states, this division shows up more clearly than in normal waking consciousness.

Psychoanalysts, I think, should be excited by this. It suggests a manifestation in dreams of two fundamental concepts of psychoanalysis Ė namely, transference and object representations Ė and holds out the possibility of identifying the neurobiological mechanisms of these phenomena.


Now, I would like to turn our attention to another strange form of dream thinking, which I call "interobjects." Interobjects are a kind of dream condensation in which the dream thoughts converge and create a new object that does not occur in waking life and could not occur in waking life. It may have a vague structure that is described as "something between an X and a Y." Allan Hobson (1988) dreamt of "a piece of hardware, something like the lock of a door or perhaps a pair of paint-frozen hinges." In dreams we accept these sorts of intermediate structures. Hobson calls them "incomplete cognitions" and Freud calls them "intermediate and composite structures." I prefer to call them interobjects. Rather than focus on what they are not (not complete condensations), I would prefer to focus on what they are (new creations derived from blends of other objects).

With interobjects, the combination of objects is not fully formed into a new object with a complete "gestalt" but rather remains incompletely fused. The created object is not easily describable in the terms of the waking world of objects. The dreamer may say, as did one of Donald Meltzerís patients, "It was something between a phonograph and a balance" (Meltzer, 1984, p. 45). It is unclear whether, in the dream itself, the objectís characteristics were vague, or merely hard to describe. An artist could probably draw a single object that was something between a phonograph and a balance, perhaps incorporating familiar parts from each object. But it is also possible that in the dream experience, the object so described was either a stable double image, or an image that shifted between two objects, or something else. An inquiry into the dreamerís experience may clarify this. Can the dreamer elaborate on how the object in the dream was "between" two objects?

Freud (1900, p. 602) seemed not to think highly of interobjects and noted how alien they were to secondary process thinking. Freud rightly notes that in our waking thinking, we avoid interobjects, lest we be thought psychotic. Yet it is quite remarkable that in our waking reports of dreams, we readily report these interobjects. As with disjunctive cognitions, the dreamer usually does not qualify his report by noting how strange it is to have perceived the interobject in his dream. He usually just reports it. Interobjects, like disjunctive cognitions, are a commonplace bizarreness of dreamlife [Footnote 4: It would be worth studying whether there are neuropathological syndromes in which interobjects are common during waking experience].

If interobjects are socially unacceptable, that does not mean that they have no use. These intermediate and compromise structures, these interobjects, may have an elementary function in human thought that has barely been explored. There are constructive aspects of extra-linguistic formations, like interobjects, that can be crucial in the formation of really new ideas that would be harder to come by using only fully-formed, secondary process formations.

They also suggest something else about cognitive processing. We tend to categorize items into discrete categories. But, at least in our dreams and perhaps in our unconscious waking life, we may find commonalities between objects that are normally kept in separate categories. The interobject reported by Donald Meltzer that I mentioned before Ė "something between a phonograph and a balance scale" -- suggests a category of objects that have circular metal platters and straight metal bars. This is not a category of objects that most of us are aware of. It is an open question whether the dream of this interobject suggests something undiscovered about mental feature detectors.

The study of disjunctive cognitions in relation to neuropathological syndromes, which I am describing, is related to what Freud (1933) called "endopsychic perception." Freud meant by this term that psychopathological symptoms like hallucinations and delusions also have a positive value for researchers, in that they may suggest how the human mind is organized and structured (Blechner, 2002). For example, delusions of being watched may be considered a projected "endopsychic perception" of the superego. Freud wrote (1933, p. 59):

"[Psychotics] have turned away from external reality, but for that very reason they know more about internal, psychical reality and can reveal a number of things to us that would otherwise be inaccessible to us. We describe one group of these patients as suffering from delusions of being observed. They complain to us that perpetually, and down to their most intimate actions, they are being molested by the observation of unknown powers Ė presumably persons Ė and that in hallucinations they hear these persons reporting the outcome of their observation: Ďnow heís going to say this, now heís dressing to go outí and so on. Observation of this sort is not yet the same thing as persecution, but it is not far from it; it presupposes that people distrust them, and expect to catch them carrying out forbidden actions for which they would be punished. How would it be if these insane people were right, if in each of us there is present in his ego an agency like this which observes and threatens to punish, and which in them has merely become sharply divided from their ego and mistakenly displaced into external reality?" (Italics mine.)

In this fascinating discussion, Freud is suggesting that the delusions of madmen reflect an insight into the organization of the mind. The delusion of being judged is an "endopsychic perception" of the morally judging aspect of the personality; the internal psychic agency, the punitive superego, is perceived as being outside the self. But the delusion itself may be evidence of the existence of the internal psychic agency.

I suggest that we expand Freudís conception of "endopsychic perception" into what I call "endoneuropsychic perception." By this I mean that certain psychological phenomena may represent an insight into the organization of the brain. The brain may be observing its own psycho-neural activity and portraying it either metaphorically or directly in dream formations. I have already suggested (Blechner, 2001, pp. 269ff.) that dreams offer us endoneuropsychic perception of many brain processes, including perception of pain, color, and time, as well as memory, reality monitoring, and other forms of meta-cognition. If we carefully examine the pattern of disjunctive cognitions and interobjects in our patientís dreams, we can provide essential data to cognitive neuroscience, and we may help formulate many new hypotheses about the structure of the human mind/brain.


Blechner, M.J. (2001). The Dream Frontier. Hillsdale, NJ: The Analytic Press.

Blechner, M.J. (2002). Hallucinations and delusions. In: Erwin, E., ed., The Freud Encyclopedia. New York: Routledge, pp. 141-142, 251-252.

Bodamer, J. (1947). Die Prosopagnosie. Archiv Psychiatrische Nervenkrankheiten, 179: 6-54.

Braun, A., Balkin, T., Wesensten, N., Gwadry, F., Carson, R., Varga, M., Baldwin, P., Belenky, G., Herscovitch, P. (1998) Dissociated pattern of activity in visual cortices and their projections during human rapid eye movement sleep. Science, 279: 91-95.

Farah, M. (1995). Dissociable systems for visual recognition: A cognitive neuropsychology approach. In: Osherson, D., ed. An Invitation to Cognitive Science. Vol. 2. Visual Cognition. Cambridge, MA: MIT Press.

Fodor, J. (1983). The Modularity of Mind. Cambridge, MA: MIT Press.

Fosshage, J. & Loew, C., eds., (1987). Dream Interpretation: A Comparative Study. Revised edition. New York: PMA Publishing.

Freud, S. (1900). The Interpretation of Dreams. Standard Edition, Vol. 4 and 5.

Freud, S. (1933). Revision of the theory of dreams. In: New Introductory Lectures on Psycho-Analysis. Standard Edition, vol. 22: 7-30.

Gorno Tempini, M., Price, C., Vandenberghe, R., Josephs, O., Kapur, N., Cappa, S., et al. (1997). Neuroimage, 5: S560.

Hobson, J.A. (1988). The Dreaming Brain. New York: Basic Books.

Hobson, J.A. & McCarley, R.W. (1977). The brain as a dream state generator: An activation-synthesis hypothesis of the dream process. The American Journal of Psychiatry, 134: 1335-1348.

McClelland, J. & Rumelhart, D. and the PDP Research Group (1986). Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Cambridge: MIT Press.

Meltzer, D. (1984). Dream-Life. Perthshire: Clunie Press.

Mesulam, M.ĖM. (1998). From sensation to cognition. Brain, 121: 1013-1052.

Perrett, D., Rolls, E., and Caan, W. (1982). Visual neurones responsive to faces in the monkey temporal cortex. Experimental Brain Research, 47: 329-342.

Tranel, D., Damasio, A., & Damasio, H. (1988). Intact recognition of facial expression, gender, and age in patients with impaired recognition of face identity. Neurology, 38: 690-696.

Mark J. Blechner, Ph.D.
145 Central Park West
New York, NY 10025
E-Mail <>

Web Editor: Paolo Migone. For suggestions or corrections, please e-mail to: