A Review of Multimodal Hallucinations

Categorization, Assessment, Theoretical Perspectives, and Clinical Recommendations


Schizophr Bull. 2021;47(1):237-248. 

In This Article

MMHs: Definition and Categorization

MMHs have been referenced in the literature by terms such as "polymodal/polysensory/intersensorial" and more,[24] reflecting a lack of consensus on how to name, categorize, and understand such phenomena. There is confusion regarding MMHs at the level of a person's range of experiences (an individual is prone to having MMHs) vs at the level of a hallucinatory episode (a particular experience can be classified as MMHs). If conceptualized at the person level, MMHs do not require temporal relatedness. Conceptualizing them at the level of a single hallucinatory episode would be more stringent but would also involve consideration of how closely together in time the hallucinations across modalities should occur to be considered part of the same hallucinatory episode (from simultaneously at some point to being in the same day or within the same psychotic episode).

Consequently, the lack of consistent specification regarding the temporal relationships that unimodal hallucinations might have with each other and whether they have to occur within a specific period to count as serial MMHs have made cross-comparisons of different studies difficult.


Given the lack of clarity in the literature, one aim of this review is to provide a categorization framework of MMHs along 3 dimensions (based on work in;[13,14,16] Figure 1. Examples of the different types of MMHs given by the possible combination of features along these dimensions can be found in Table 1. For the implications of such framework for clinical and research practice see "Clinical Implications" section.

Figure 1.

Categorization of multimodal hallucinations (level of a hallucinatory episode) across 3 dimensions: time, relatedness, and congruence. The arrows show how the different categories/levels can be combined to give rise to different types of multimodal hallucinations.

The first dimension considers whether the hallucinations are occurring on the same temporal scale across multiple sensory modalities. If they co-occur, they are categorized as "simultaneous MMHs." If they occur in one sensory modality at a time (unimodal) but in a sequential manner over time (with delays ranging from minutes, days, and more), they are categorized as "serial MMHs." By conceptualizing such experiences at the person level instead, a clearer dimensionality emerges: one would consider the proportion of hallucinatory experiences that are in multiple modalities simultaneously vs in different modalities at different times.

The second dimension looks at whether the MMHs are experienced as being "from the same source or entity,"[14,16] but how to understand this concept of a common source is underspecified in the literature. A case study[25] of a patient seeing and hearing the voice of a human figure shows how the common source can be based on perceptions of the same entity in different sensory modalities. However, hallucinations might also be of distinct but semantically related entities (eg, having a religious vision and subsequently hearing the voice of God). Importantly, considering relatedness in a dimensional rather than dichotomous way suggests 3 important points along this dimension: (1) cases of maximal relatedness in which MMHs represent the same entity; (2) cases of moderate relatedness in which MMHs represent distinct but meaningfully related entities; and (3) cases in which hallucinations across different modalities represent completely unrelated entities.

Finally, the third categorization level concerns whether the combination of hallucinations across modalities is contextually coherent (congruent MMHs) or not (incongruent MMHs). One limitation is that special combinations of the dimensions might give rise to definitional issues of MMHs. Would a (temporal) sequence of (conceptually) unrelated hallucinations occurring within a short time frame be considered multimodal? Perhaps yes.

It is important to highlight that this classification system is a clinical heuristic that would need further validation. Crucially, service users experiencing MMHs and unimodal hallucinations should be consulted in order to validate the current framework and available measures (discussed later) to ensure that one does not impose a classificatory system that limits their understanding of such experiences.