Note
Unconscious vision in action

https://doi.org/10.1016/j.neuropsychologia.2007.09.005Get rights and content

Abstract

The dorsal visual processing stream has been suggested to be involved with the unconscious processing of visual information for visually guided actions. In this study, transcranial magnetic stimulation (TMS) was used to gate input into the dorsal visual processing stream by disrupting primary visual cortex (V1) function. Despite restricting geniculostriate processing contributions in the dorsal stream, consistent effects on reaching performance from unconscious visual events were nonetheless measured. These results suggest a functionally intact, albeit unconscious, projection to the dorsal pathway that bypasses V1 and implicates a functional input into the dorsal stream from the superior colliculus.

Introduction

In our everyday lives, we infrequently reach out and grasp an object that we have no intention of using. For example, when preparing a meal, we may suddenly be surprised to find a tin can opener in our hands when trying to open a glass jar. Such instances suggest that our visual and motor systems may be independently linked from conscious perception. The mechanisms underlying these visual ‘zombie’ processes have been under intense investigation and have provided some insights into the neural basis for unconscious vision (Milner & Goodale, 1995). In particular, the dorsal visual processing stream has been suggested to be involved with the coding of visual object locations (Ungerleider & Mishkin, 1982), and more recently, for visually guided actions (Goodale & Milner, 1992).

The primary evidence for dorsal stream involvement in vision for action comes from demonstrations of preserved reaching and motor control abilities in a patient with apperceptive visual object agnosia after bilateral ventral, but not dorsal visual cortex damage (Goodale, Milner, Jacobson, & Carey, 1991; James, Culham, Humphrey, Milner, & Goodale, 2003). In addition to dorsal stream projections from the visual cortex, several anatomical studies have also demonstrated a small proportion of direct retinal projections into the superior colliculus, which in turn sends projections through the pulvinar nucleus of the thalamus to the posterior parietal cortex of the dorsal stream (Kaas & Huerta, 1988; Robinson & McClurkin, 1989). Thus, influences on visually guided actions may be a consequence of dorsal stream projections from both the retinogeniculostriate and retinotectal pathways. In this study, I tested the hypothesis that information processing within the retinotectal pathways alone is sufficient to influence visually guided actions in humans.

To assess this hypothesis, a remote distractor paradigm was adapted to be used in a visually guided reaching task. In this paradigm, a visual distractor typically produces saccadic eye movement onset delays to a simultaneously presented target (Rafal, Smith, Krantz, Cohen, & Brennan, 1990; Ro, Shelton, Lee, & Chang, 2004; Walker, Deubel, Schneider, & Findlay, 1997). However, when simple button press reaction times to the onset of two simultaneously presented targets are measured, responses are faster when an additional target is presented, a phenomenon referred to as the redundant target effect (Marzi, Tassinari, Aglioti, & Lutzemberger, 1986; Miller, 1982). Interestingly, both the onset delays in saccades from remote distractors and the decreases in response times for button presses from redundant targets have been measured without awareness of the additional stimulus (see Rafal et al., 1990, Ro et al., 2004 for saccades; and Marzi et al., 1996; Savazzi & Marzi, 2002; Tomaiuolo, Ptito, Marzi, Paus, & Ptito, 1997 for button presses). In addition to assessing any effects of unconscious stimuli on reaching performance, this study also examined whether delays in reaching onset, as with saccades, or whether faster reaching onset latencies, as with manual button presses, are induced from an additional visual stimulus presented along with a target.

Section snippets

Methods

Transcranial magnetic stimulation (TMS) over the primary visual cortex was used to induce a transient visual suppression and to limit visual information processing to non-geniculostriate visual pathways. Prior to commencing the main experiment and after informed consent, a visual cortex localization task was performed in each of the six participants (mean age = 25.8; 2 males). TMS was conducted using a Cadwell Laboratories (Kennewick, WA) MES-10 stimulator (2.2T maximum output) connected to a 9 cm

Results

On the no TMS control trials, not surprisingly, participants were highly accurate (94.2%) at detecting the irrelevant visual stimulus and very rarely (0.8%) made a false alarm. In contrast, when TMS was applied over the primary visual cortex, participants made significantly more false alarms (2.1%) on the irrelevant stimulus-absent trials as compared to the no TMS trials (two-tailed t5 = 3.50, p = .017). These higher rates of false alarms may be due to the perception of induced phosphenes from TMS

Discussion

In this study, single-pulse TMS was applied over the visual cortex to disrupt the processing of a centrally presented visual stimulus. When the TMS rendered the participants blind to this centrally presented visual stimulus, influences from this unconscious event were still measured on the reaching reaction times. Thus, these results demonstrate a consistent and reliable effect of unconscious visual stimuli on visually guided actions, as has also been shown in visual masking studies (Binsted,

Acknowledgements

I thank Stephenie Harrison and Jennifer Boyer for assistance with collecting some of the data. The author has no competing financial interests.

References (30)

  • I.M. Blythe et al.

    Residual vision in patients with retrogeniculate lesions of the visual pathways

    Brain

    (1987)
  • E. Corthout et al.

    Timing of activity in early visual cortex as revealed by transcranial magnetic stimulation

    Neuroreport

    (1999)
  • A. Cowey et al.

    Magnetically induced phosphenes in sighted, blind and blindsighted observers

    Neuroreport

    (2000)
  • R. Fendrich et al.

    Residual vision in a scotoma: Implications for blindsight

    Science

    (1992)
  • M.A. Goodale et al.

    A neurological dissociation between perceiving objects and grasping them

    Nature

    (1991)
  • Cited by (35)

    • Is the primary visual cortex necessary for blindsight-like behavior? Review of transcranial magnetic stimulation studies in neurologically healthy individuals

      2021, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      The authors did not observe a similar effect when the participants reported the location of the target with a manual button-press, based on which they argued that the effect was mediated by visual pathways through the SC. Later, using a similar distractor paradigm, Ro (2008) observed that it was possible to influence reaction times when a manual reaching task was used. Both of these studies (Ro et al., 2004; Ro, 2008) employed a binary yes/no visibility rating.

    • Dissociations of conscious and unconscious perception in TMS-induced blindsight

      2019, Neuropsychologia
      Citation Excerpt :

      Investigations of blindsight, a neurological disorder characterized by visual discrimination without conscious perception after damage to V1 (Weiskrantz, 2009; Weiskrantz et al., 1974), have suggested both subcortical (Dodds et al., 2002; Pöppel et al., 1973; Rafal et al., 1990) as well as cortical (Fendrich et al., 1992; Schmid et al., 2010) mechanisms for unconscious vision. As with blindsight patients, TMS over V1 at approximately 100 ms after visual stimulus onset reliably causes blindsight behavior, whereby normal observers can correctly discriminate the attributes of visual stimuli despite their inability to detect them (Allen et al., 2014; Boyer et al., 2005; Christensen et al., 2008; Jolij and Lamme, 2005; Railo et al., 2012; Ro, 2008; Ro et al., 2004). The later time window of disruption that produces TMS-induced blindsight may suggest that unconscious vision relies on earlier feedforward stages of visual processing in V1, such that disrupting later feedback activity will affect only visual awareness (Lamme, 2001; Lamme et al., 2000).

    • The chronometry of visual perception: Review of occipital TMS masking studies

      2014, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      Even in the absence of conscious perception these patients can relatively accurately process some information about stimuli presented in the blind areas, such as motion direction or the location of stimuli. Ro et al. (2004) showed in normal participants that, when TMS pulses to early visual cortex successfully masked visual stimuli, these stimuli could still affect saccades (see Christensen et al. (2008) and Ro (2008) for effects on reaching movements). Jolij and Lamme (2005) showed that TMS suppression of the visibility of emoticons (subjects could not localize the emotional emoticon in an array of neutral emoticons) did not necessarily abolish processing of the emotional content (subjects could still indicate whether the emoticon was ‘sad’ or ‘happy’), demonstrating ‘affective blindsight’ in normal observers.

    View all citing articles on Scopus
    View full text