The Unity of Consciousness

At a moment, sights, sounds, bodily feelings, thoughts, and intentions seem present to one subject within one field. This unity of consciousness is not one simple property. Experiences may belong to one subject without being mutually accessible, represent features in one scene without sharing a modality, or be phenomenally co-present without supporting one action. Distinguishing these unities is necessary before asking whether the brain constructs them or a simple self supplies them.

Unity matters because theories of the self often use it as evidence. Substance dualism appeals to an indivisible subject; bundle and functional theories seek relations that bind distributed states. Split-brain and dissociation cases test those relations, but no behavioural result counts subjects until a criterion of unity is chosen.

Subject unity

Experiences exhibit subject unity when they are all experiences of one subject. Your visual experience and pain are both yours even if they concern unrelated objects. This is a metaphysical ownership claim.

A simple-substance theory makes subject unity primitive: one bearer has many states. A functional theory identifies a subject through integrated access, memory, and control. A no-self view may treat ownership as conventional shorthand for causally connected processes.

Subject unity cannot be inferred simply from spatial location. Two subjects might share one body, and one distributed subject might extend beyond a conventional boundary. Nor does qualitative similarity unite states; matching experiences can belong to different people.

Phenomenal unity

Experiences are phenomenally unified or co-conscious when they are experienced together as parts of one total phenomenal field. Seeing the screen and hearing music are not merely simultaneous; there is something it is like to undergo them jointly.

The unity thesis says that, at any time, all a subject's conscious experiences are co-conscious. Possible counterexamples include peripheral or unattended experience, dissociation, and divided attention. Whether these are failures of phenomenality, access, or report is contested.

Phenomenal unity appears stronger than common ownership. One subject might have two streams inaccessible to one another, as some interpretations of split-brain cases suggest. If so, subject unity does not guarantee field unity.

Representational unity

Perception presents objects and properties within organised scenes. Colour is bound to one shape rather than another; sounds and lip movements are assigned to one event; locations are integrated across modalities. This is representational unity.

The binding problem asks how distributed processing yields correct combinations. Mere synchrony or co-location may be insufficient; attention, recurrent processing, object files, and common causal sources are candidate mechanisms.

Representational unity can occur without phenomenal unity if unconscious systems bind features. Conversely, a confused or fragmented scene may still be experienced within one field. The two are empirically related but conceptually distinct.

Access unity

States possess access unity when their contents are jointly available to reasoning, working memory, report, and action. A global-workspace theory makes this broadcast central to consciousness.

Access can bind information across modalities and coordinate one response. It may be narrower than phenomenal consciousness if experience overflows report, or broader if unconscious information affects decisions.

One subject's control system can also contain partially isolated modules. Unity is then graded by bandwidth, recurrence, and which consumers share the information rather than an all-or-nothing switch.

Agency unity

An agent ordinarily forms plans, resolves conflicts, and controls one body as a coordinated whole. Agency unity concerns integration among intentions, evaluations, and actions.

It can fail while consciousness remains unified. Akrasia, dissociated control, alien- hand phenomena, and compulsive action reveal conflict among systems. Conversely, coordinated behaviour can emerge from multiple processes without one phenomenal field.

Agency unity is especially relevant to personhood and responsibility. It should not be used as a direct measure of how many phenomenal subjects exist.

The subsumptive view

One account says unified experiences are all parts or aspects of a single total experience. At each moment, local experiences are subsumed by a maximal phenomenal state.

This captures the apparent field structure and explains pairwise co-consciousness through common inclusion. It raises questions about boundaries and overlap. Could two total experiences share a component? What fixes the temporal grain of a total state? Does subsumption explain unity or restate it mereologically?

The view is compatible with physical realisation if a neural integration process corresponds to the total state. It does not identify which process is sufficient.

The relation-first view

Another account treats co-consciousness as a primitive or analysable relation among experiences. A set of states belongs to one field when connected by the right reciprocal access, integration, or phenomenal relation.

This fits bundle theories: a subject may be constituted by a maximally unified set rather than unity resulting from a prior owner. The relation must avoid transitivity problems. If $A$ is co-conscious with $B$ and $B$ with $C$, must $A$ be co-conscious with $C$? Partial and overlapping unity may violate simple equivalence-class models.

Empirical cases can motivate graded graphs rather than discrete bundles, but metaphysical subject count then becomes correspondingly indeterminate.

The simple-subject view

A simple self supplies unity because every experience modifies one indivisible subject. This reverses the explanatory order: co-consciousness follows from ownership rather than ownership from integration.

The view handles unity cleanly but makes division difficult. If access and control split, does the simple subject remain one? Evidence cannot directly reveal simplicity, and a primitive bearer may add no prediction beyond the integrated processes it owns.

The substance-dualism chapter develops the associated divisibility argument and its limits.

Neural integration

Candidate mechanisms include recurrent interaction, thalamocortical coordination, global broadcast, temporal synchrony, and integrated causal structure. Each captures a different unity: information sharing, access, feature binding, or irreducibility.

A correlation between integration and unified report does not prove that integration constitutes phenomenal unity. Yet mechanistic dissociations can rule out simple proposals. If one kind of binding survives while phenomenal or access unity fails, the mechanisms should be separated.

The empirical theories of consciousness compare these proposals as theories of consciousness. Here they are assessed only for the unities they explain.

Is unity necessary?

Some theories treat consciousness as necessarily unified: a conscious state must be present to a subject within a field. Others permit isolated phenomenal fragments or multiple overlapping streams.

Necessity may arise from concept rather than evidence. If a subject is defined as the bearer of a maximal unified field, then unowned fragments are excluded by terminology. The substantive question is whether phenomenal occurrences require integration beyond their local character.

Fragmented cases are hard to report precisely because report requires access unity. Absence of reports about fragmentation is therefore weak evidence that fragmentation never occurs.

Unity and time

Consciousness also appears unified across a short temporal window. Melody, motion, and speech are experienced as temporally extended patterns, not isolated instants. This diachronic unity may depend on retention, anticipation, memory, and integration over hundreds of milliseconds or longer.

Diachronic experiential integration is not personal identity through years. It explains how one present episode includes just-past phases. The metaphysics of persistence remains in Persistence, while temporal experience is treated in Temporal Passage and Experience.

Assessment

No one empirical measure tracks every unity. The philosophical work is to specify which unity a theory treats as basic and how the others arise. Split-brain cases are valuable because they pull the relations apart; they do not carry a ready-made count of subjects.

Selected references

• Bayne, Tim. The Unity of Consciousness (2010).
• Brook, Andrew and Paul Raymont. The Unity of Consciousness (2014).
• Dainton, Barry. Stream of Consciousness (2000).
• Hurley, Susan. Consciousness in Action (1998).