Split Brains and Counting Subjects

In some patients, the corpus callosum connecting the cerebral hemispheres has been severed to treat severe epilepsy. Carefully arranged tasks can then deliver information to one hemisphere while preventing ordinary sharing with the other. Responses may diverge: one hand selects an object the speaking hemisphere says it did not see, or the two hands pursue conflicting tasks. Split-brain cases ask whether one person now contains two conscious subjects, one divided subject, or one subject with partially isolated systems.

The data do not announce a subject count. Experiments measure report, discrimination, memory, and control under unusual conditions. Moving from those capacities to one or two minds requires a theory of unity, and different unities may divide to different degrees.

The basic anatomy and method

The cerebral hemispheres normally exchange information through commissural pathways, especially the corpus callosum. Commissurotomy reduces this transfer without producing two detached half-brains: subcortical connections, common sensory input, one body, and environmental coordination remain.

Visual stimuli can be lateralised briefly because each visual hemifield projects primarily to the opposite hemisphere. Language is often left-lateralised, so a stimulus presented to the left visual field may guide the left hand while the speaking system denies seeing it.

This experimental arrangement is delicate. Eye movements, stimulus duration, cueing, cross-cueing, individual variation, and recovery matter. Laboratory dissociation does not imply complete division in ordinary life.

What is divided?

Split-brain evidence can show division in:

• perceptual access to lateralised stimuli;
• verbal report;
• memory and recognition;
• motor planning and hand control;
• metacognitive access;
• integration across response systems.

It does not directly show division in phenomenal consciousness. A non-speaking hemisphere's accurate selection may reflect conscious perception, unconscious processing, or a degree of awareness difficult to report.

The inferential problem resembles blindsight but adds reciprocal organisation and possible independent control. Multiple measures are needed before identifying a second stream.

The two-subject view

On a two-subject interpretation, commissurotomy creates or reveals one conscious subject associated with each hemisphere. Each has distinct perceptual access, memory, and control, though only one ordinarily speaks.

The view explains simultaneous incompatible responses and the apparent private information of each hemisphere. If subjecthood follows a maximal integrated conscious field, two fields suggest two subjects.

Objections stress continued unity. Patients generally present as one person in daily life, share one body and environment, and retain subcortical integration. Hemisphere systems are not equally capable across domains. Counting every partially autonomous processor as a subject risks proliferation within ordinary brains.

The one-subject view

A one-subject interpretation holds that the person remains one conscious subject with divided access and response channels. One subject can have information unavailable to one mode of report, just as ordinary cognition contains unconscious and dissociated processes.

Biological continuity, unified life, and ordinary behaviour support this view. It avoids the abrupt creation of a second person by surgery.

The difficulty is explaining concurrent incompatible intentions or experiences. If one subject consciously sees both stimuli, why can the speaking system not access one? If the subject alternates or has fragmented consciousness, the unity claim becomes thinner than ordinary one-subject experience.

Partial unity

A third view rejects a binary count. Consciousness may be unified along some dimensions and divided along others. Two perceptual fields can share arousal, bodily affect, and long-term goals while lacking direct informational access.

Represent this as a network rather than an integer:

graph LR
  L[Left-hemisphere processes] <-->|subcortical and environmental coupling| R[Right-hemisphere processes]
  L --> B[One body and life]
  R --> B

Partial-unity views fit graded integration and avoid forcing complex organisation into exactly one or two. They make subjecthood itself graded or indeterminate, which some regard as a cost. Moral and legal practice often requires discrete persons even when underlying unity is not discrete.

The switch model

Perhaps one subject alternates between hemisphere-dominated streams depending on task and attention. At a given moment one system controls report while another influences action, without simultaneous full subjects.

This model can explain variability and ordinary unity. It must distinguish rapid alternation from simultaneous dissociation experimentally. Conflicting bimanual action and concurrent task performance pressure a simple serial switch.

The broader lesson is temporal resolution. Subject count can vary with the window over which integration is assessed; momentary fragmentation may coexist with longer-term coordination.

The interpreter hypothesis

Research on split-brain report suggests the speaking hemisphere constructs explanations for actions initiated by information it lacks. This interpreter produces coherent narratives from partial evidence.

The finding bears on self-knowledge: fluent explanation can be sincere and causally inaccurate. It does not show that narrative is always false or that only one hemisphere has consciousness.

It also cautions against equating a single verbal voice with a single subject. Speech may integrate and present one account while non-verbal systems retain information and control unavailable to it.

Counting criteria

Possible criteria for counting subjects include:

No criterion is neutral. The case tests theories by forcing their criteria to yield a verdict and absorb its consequences.

Fission versus split consciousness

Philosophical fission cases divide a person's psychology into two future continuers. They concern numerical identity across time. Actual split-brain cases primarily concern simultaneous organisation within one ongoing organism.

The issues interact but should not be collapsed. Evidence of two current subjects would not by itself settle which is identical with the pre-surgical person. Conversely, identity could be indeterminate while current consciousness remains largely unified.

The fission and persistence questions remain in Persons and Minds in the Inventory and Persistence.

Dissociative identity

Dissociative identity disorder involves discontinuities in memory, agency, identity, and self-experience. It is sometimes compared with split brains, but the mechanisms and evidence differ. Psychological dissociation does not involve surgical separation of hemispheres, and popular depictions exaggerate discrete, internally uniform persons.

The comparison is useful only at the level of criteria: how much separation in memory, control, and self-representation warrants speaking of multiple selves or subjects? It should not be used to infer one condition's structure from the other.

Clinical categories serve treatment and communication as well as metaphysical counting. Respect for persons does not require a settled theory of how many phenomenal fields a condition contains.

Conjoined and distributed systems

Conjoined twins and experimentally coupled agents provide a reverse comparison: two organisms or established subjects can share sensory information and control. Increased integration need not merge subjects automatically.

This pressures purely functional criteria. If information sharing creates unity, how much is enough? If prior biological individuality blocks merger, subjecthood depends on history as well as current organisation.

Distributed artificial systems raise similar issues. Components can share memory and control while retaining partly independent processing. Split-brain cases show that architecture alone needs a principled grain and temporal scale before yielding a count.

Moral and practical identity

Legal and interpersonal practices usually treat a split-brain patient as one person. This is supported by one body, biography, social identity, and coordinated life. A metaphysical possibility of partially divided consciousness need not force immediate revision of personhood.

If genuinely independent centres of welfare or agency existed, however, moral attention might need to track each. Subject count and person count can diverge: one legal person could contain multiple centres of experience, or one subject might lack full personhood.

The next page separates these normative criteria explicitly.

Assessment

Split-brain research establishes dissociations in access and control, not a simple verdict of "two minds." The evidence is compatible with one divided subject, two asymmetrical subjects, switching, or partial unity depending on the chosen account of consciousness and subjecthood.

This underdetermination is productive. A theory that says subjects are simple substances, global workspaces, maximal phenomenal fields, organisms, or narrative agents must say which observed divisions matter and why. Counting is the output of that theory, not a datum read directly from a hand response.

Selected references

• Bayne, Tim. The Unity of Consciousness (2010).
• Gazzaniga, Michael S. Who's in Charge? (2011).
• Nagel, Thomas. "Brain Bisection and the Unity of Consciousness" (1971).
• Schechter, Elizabeth. Self-Consciousness and Split Brains (2018).