In Chapter 11, Dennett starts investigating the implications of the multiple drafts model, revisiting the reporting problem introduced in the previous chapter. Chapter 11 is called “dismantling the witness protection program”. Titled “Dismantling the Witness Protection Program,” the chapter directly addresses the central concept of a “witness” that makes sense of the information in our minds.

Blindsight makes us unconscious of specific stimuli like elements of vision, but conscious of others. Sometimes, for various reasons, we lose sight of one spot. For example, in right hemianopia people have complete blindness solely in the right hemifield. We all have blind spots in our vision due to each eye’s limited viewing angle. Our brains usually don’t need to address this because our eyes compensate for each other’s blind spots. Nothing in our brain reports this absence of information. However, when an unusual blind spot occurs, missing information would be expected and reported. Despite this, in experiments, blindsight subjects can somehow sense when there’s light in their field of vision. But why is this important for understanding consciousness? We’ll explore the answer.

What about people with hysterical blindness and the evidence for blindness in general? If we ignore the experimental evidence and focus on heterophenomenology—what people report about their experiences—it’s clear that even with the available evidence, we can’t fully dismiss the fact that hysterically blind individuals cannot see. Dennett then introduces another case related to blindness and consciousness. Some blind people claim that they can sense light, even though they assert they are not conscious of it. From a heterophenomenological perspective, we cannot deny the existence of their experience, but we can still question what they are actually experiencing. In this case, their experience could simply be guesswork, and blind subjects tend to improve their guessing abilities in this type of experiment. Their guesses could stem from something inherent in the experiment’s design, perhaps an unintentional signal from the experimenter. However, guessing alone is not sufficient for us to be conscious of something. Consciousness involves an “aboutness” – we must be able to define what we are conscious of and understand what is it about.

When we search for something that we haven’t seen yet, like the thimble in a ‘hide the thimble’ game, we often find that we can’t spot it even when it’s right in front of us. Then, after we do notice it and examine it, it becomes “in a position where it can be reported” as Dennett explains. Training enhances our ability to identify certain effects. For instance, blind people’s guesses improve with practice just as the abilities of sommeliers do. With more exposure and experience, we can become conscious of things we weren’t previously conscious of. Our experience isn’t fundamentally different from that of blindsight. Blindsight subjects can also strengthen the link to their experiences as their guesses improve. This introduces another element in the understanding of consciousness: qualia – how things, look, smell, or sound to us. 

Dennett introduces all these cases to discuss the next Cartesian theatre illusion: figment filling. This concept assumes that our brain fills in details about colouring in an image, missing information, or the blind spot in our vision. Many thinkers adopt this concept, but Dennett refutes it for a clear reason: there wouldn’t be anyone to report the missing part if unless we assume the existence of a central mechanism to decide to fill it in. Without such a mechanism the missing information would be simply ignored. This challenge is important because it introduces another feature of consciousness: discontinuity. 

“The absence of representation is not the same as the presentation of absence. And [vice versa]” Dennett argues, introducing the idea of virtual presence. How can this concept be applied to our information systems, like libraries? If we were to view how we retrieve information from thousands of stored pieces of information, how would we do it if our system operated in a decentralised mode, observing the external world without noticing what is missing? 

How can the multiple agents in our brains decide what information to store in the library? Dennett suggests that the information would be stored “at no cost”, and later retrieved by a system resembling a library loan system where information is “virtually present”. However, this loan system is distributed across millions of sentries – agents responsible for detecting anything novel in our sight, or our senses. These agents continuously log information about novelty around. However, they don’t waste effort in filling figments in everything our vision detects.

We don’t get notified by alarms when changes occur during saccades – rapid eye movements – something Dennett experienced during an eye-tracking experiment when the screen changed only while his eyes were moving. This phenomenon is known as “saccadic suppression” suggests similar constraints on what information what information our library system can store, as they define the boundaries of what is present for us.

Dennett explored this concept to engage in a debate with Otto, his virtual critic. Otto argued that by denying our processing of the details of the wallpaper (featuring a repeated Marylin image), Dennett was also denying his experience. Otto pointed out that, despite this denial, he does see and recognize the details of the Marilyn image. 

Dennett denies the existence of any specific module in the brain – or mind – dedicated to “figment filling”, whether for wallpapers or elsewhere.  He argues that the gaps we perceive aren’t filled by a mental process but are instead judgements made by our brains based on the information our vision detects. In the example Dennett discusses with Otto, where high-resolution Marilyn images are involved, you might recognize them as Marilyn if you focus, but you might not if you don’t. The brain receives information and makes judgments about it even when you aren’t consciously aware. These judgments could be correct or incorrect. This process explains why you can recognize Marilyns: it’s not because there’s a specific process that fills in vague details but because your brain evaluates and interprets the available information even when you’re not aware.

Similarly, imagining something might lead us to think that there is a central meaner or a Cartesian theatre where images are presented or gaps are filled. In Otto and Dennett’s dialogue, another example is the image of a pink ring. Dennett asks Otto to consider a scenario where something is presented in the brain in a sort of internal language, like mentalese (a hypothesis suggesting an internal language of thoughts in the mind). We could envision the pink ring being presented in this internal language. However, Dennett argues that there is no one to present that to, no central meaner to receive this image. Instead, multiple processes are involved in the brain to handle this information. Also, no single discerner or central entity is responsible for receiving and interpreting this information; rather it is processed in various areas throughout the brain. Some of these processes might lead to the formation of sentences, but the content itself is not language-like. Thus, there is no need for a central meaner or a language of thought like mentalese. Dennett addresses Otto’s suggestion and responds:

“Otto: So presentments [he means the presentments that appear in the brain] are like speech acts except that there’s no actor and no speech.

Dennett: Well. Yes. What there is, really, is just various events of content-fixation occurring in various times in the brain. These are nobody’s speech acts, and hence they don’t have to be in a language, but they are rather like speech acts; they have content, and they do have the effect of informing various processes with this content. We considered more detailed versions of this in chapters 5-10. Some of these content-fixations have further effects, which eventually lead to the utterance of sentences – in natural language – either public or merely internal. And so a heterophenomenological text gets created. When it’s interpreted, the benign illusion is created of there being an Author. This is sufficient to produce heterophenomenology

Otto: But what about the actual phenomenology?

Dennett: There’s no such thing. Recall our discussion of the interpretation of fiction. When we come across a novel that is loosely veiled autobiography, we find we can map the fictional events onto many of the real events in the author’s life, so in a strained sense the novel is about those real events. The author may not realize this at all, but in this strained sense it is true; those events are what the text is about because those events are the real events that explain why this text was created”

When we read a novel, we form judgements, similar to how we judge the colour of the wallpaper. These judgements arise from our existing knowledge and experiences: text provides the basis for our understanding and interpretation in the case of novels, while colours serve as the reference point for our assessments of paintings. 

Consciousness isn’t a plenum. It’s “gappy and sparse it doesn’t contain half of what people think is there”. Here we may understand why Dennett introduced the concept of heterophenomenology. One reason is to acknowledge the limitations in capturing someone’s phenomenology or what they can actually report about it. Psychotherapists may agree with Dennett on interpreting a sentence, and how to handle subjects’ reports. This perspective also helps clarify Dennett’s earlier point that consciousness isn’t continuous.