Consciousness research has a problem. How can we investigate the hidden mental processes that make up our internal world? Subjective experience is hard to document scientifically. It is not time-locked to sensory stimulus, which can be controlled, or behavior, which can be observed.
In human studies, the simplest approach to gain insight into subjective experience is to ask. Until recently, this approach was only routinely practiced in the realm of psychology. As human brain imaging and recording technologies have advanced, researchers are developing methods to link subjective experiences with brain activity.
Seeking neural correlates of consciousness
One major aim of consciousness research is to identify ensembles of neurons that exhibit electrical activity in tight correlation with subjective experiences. From that foundation, it is believed, the harder question of what the physical nature of consciousness is can be addressed.
Decision making is one of the most studied internal processes. In the controlled setting of a well-designed experiment, decisions are kept simple, and any sensory stimuli are vetted so that the collected data stands a chance of being interpreted accurately. For example, in a recent study by Chiara F. Tagliabue and Domenica Veniero et al., subjects were asked to decide whether a particular visual stimulus was brighter or darker than its background. Contrast was varied between three different levels, predetermined to result in a 25% (low), 50% (intermediate), or 75% (high) detection rate. That means sometimes the contrast was so slight, subjects could only see the visual stimulus 25% of the time.
This is similar to a commonly given visual field test that you may have taken. Leaning into eyepieces that fit like goggles to block ambient light, you are asked to click a button whenever you see a black dot on a bright screen. The location, intensity, and duration of the dot stimuli are varied to test for blind spots within your visual field. During the test, you are actively deciding whether or not you are seeing a dot.
To make things more interesting, Tagliabue and Veniero followed the stimulus with two questions. First, subjects must make their decision: “Was the stimulus lighter or darker than the background?” Second, subjects must rate their experience of the stimulus: “How clearly did you see the stimulus?” Subjects were asked to select one of four options on a Perceptual Awareness Scale: 0(no experience), 1 (brief glimpse), 2 (almost clear) or 3 (clear). It is with this follow-up question that the researchers attempted to investigate individual, subjective experiences.
To search for the neural correlates of subjective experiences, a funky net of electrodes were noninvasively attached to the subject’s scalp prior to testing. During the decision making test, neural activity was simultaneously recorded using the net, an EEG (electroencephalogram). EEG has the following limitations. First, it can only record electrical activity from the outermost layers of the brain — the cortex. Second, EEG electrodes can only pick up the simultaneous electrical activity of large neuronal ensembles, which are measured as Local Field Potentials. To interpret results between individuals and experiments, electrodes are always attached in standardized locations.
Tagliabue and Veniero went into this study knowing that a specific type of EEG signal, called the centro-parietal positivity (CPP), was important for decision making. The CPP is named for its location in the center of the parietal lobe of the cortex, and for its characteristic late positive EEG signal, thought to be indicative of evidence accumulation.
It had previously been shown that the CPP signal builds as an individual is considering sensory evidence. The signal builds faster when sensory evidence is strong (so in this case, high contrast) and peaks at the decision point. CPP signals are not specific to any particular type of sensory stimulus or behavioral output.
To better understand whether or not the CPP signal may underlie the internal processing that occurs during decision making, Tagliabue and Veniero analyzed the CPP signal from the EEG data, and determined whether or not it co-varied with the strength of visual stimulus (objective, physical stimulus) or the subject’s selected value from the Perceptual Awareness Scale (subjective).
As you might expect, the CPP signal did increase when the intensity of the stimulus was increased — a high intensity stimulus is also more likely to be perceived clearly by the subject. The results were that the CPP voltage always started to increase about 200 ms after the visual stimulus and peaked at around 400–600 ms, reflecting the lag between the exposure to the stimulus and the perceptual awareness of it.
To further tease apart whether the CPP was representative of the physical stimulus or subjective processing of that stimulus, Tagliabue and Veniero considered one variable at a time. When comparing trials where the stimulus intensity could be equated, the CPP voltage scaled with the subject’s perceptual awareness value (PAS in the figure below). Stimuli that were perceived clearly generated a higher CPP voltage peak than the stimuli that were not perceived clearly, or not detected at all.
When comparing trials where the perceptual awareness value could be equated, the CPP voltage did not change with the strength of the visual stimulus. This suggests that the CPP signal represents the subjective clarity of the stimulus, and not the objective strength of the stimulus.
Interestingly, the CPP signal was present even in catch trials, where no visual stimulus was presented. Perhaps this represents the subject’s expectation of a stimulus. When experiencing anticipation, the brain becomes primed to detect incoming stimuli. To use a video camera as an analogy, the CPP signal is like pressing the record button in anticipation of an interesting scene. I bet the arousal system described in a recent post plays a role in generating the catch trial CPP signal.
Are we reliable narrators of our own experience?
Our study suggests a remarkable degree of reliance that can be placed on observers’ subjective impressions to gain access to core internal decision quantities.From Tagliabue and Veniero et al. (2019)
To understand subjective experience, we have to rely on human studies, since human beings are the only animals who can tell us their subjective experiences. But can people be trusted to accurately report their experiences?
There are some significant pitfalls about self-reporting. First, actively reporting subjective experience as it is happening changes the experience.
Second, accuracy decreases with time. People rapidly assimilate experiences into a larger narrative. If a subject must wait until a post-testing interview, what they say will no longer reflect their experience in the moment.
Tagliabue and Veniero addressed this problem by asking that subjects pick from one of four coded responses immediately after each visual stimulus was presented. Unfortunately, that is limited to a controlled lab setting. Can this approach be modified to investigate the neural correlates of the complicated, noisy, fragmented subjective experiences we have in real life settings? I am not so sure, but it may be worth a try.
Let’s say we believe the conclusions of the paper by Tagliabue and Veniero, and are ready to declare that the centro-parietal positivity is the neural correlate of perceptual awareness. So what? Are you satisfied? Do you feel like you have gained insight into what it means to have subjective experience? What would the next steps be?
The purpose of these questions is not to rip apart Tagliabue and Veniero and their findings. I think their approach was solid and their results were interesting. The purpose is to question whether studies like these can inform us about the nature of consciousness. I have my own thoughts, but will leave these as open questions for now.