You might know the word “knish”—you’ve heard or read it—but not know what it refers to. In that case, you know the word’s sound or spelling, but you don’t really have the concept. But if you truly have the concept KNISH, it wouldn’t make sense to say you have no idea what a knish is, because some understanding is part of having the concept?

The Chinese Room illustrates the symbol-grounding problem from lecture: following symbol rules could produce an appropriate answer without the person following those rules understanding the language. The Chinese symbols still have meanings for Chinese speakers, but the problem is that JS does not know those meanings.

So modal representations (use sensory information) and amodal representations (uses abstract symbols). But even if a brain pattern is tied to vision or movement, how does that make the pattern mean chair or number five? In lecture we talked about embodied concepts being as one possible answer, while this chapter questions whether embodiment alone fully explains meaning

People asked to list watermelon features without special instructions responded more like people told to imagine a watermelon than people doing word association. That pattern supports the possibility that ordinary thought draws on perceptual representations. It is evidence for modal concepts,but are the represeentations really necessary for understanding??

people with different color vision may conceptualize color differently, and handedness may influence preferences for left- or right-sided objects. The Fribbles study tests whether right- and left-handers differ in the way they connect sides of space with positive or negative traits. So this furthers the argument of the affect is not just an automatic consequence of having a body

This describes what concepts do, but leaves open how they are represented. The lecture contrasts the standard view—concepts as amodal symbols in a semantic network—with the embodied view, where concepts use sensory or motor information. So embodied concepts still needs to explain how people can use concepts consistently and communicate about the same things.

The color example in lecture helps make this claim concrete: someone with different cone cells may experience color differently, and someone who cannot see may not form color concepts in the same way. That supports the idea that our bodies shape the world as we experience it. It doesn’t necessarily mean there is no world outside the mind; shared biology and culture can restrict how people experience it

Our bodily and sensory experiences shape how we think about things. The slides describe concepts as grounded in analog, sensory-based representations, rather than only in abstract verbal definitions. For example, what someone knows about a “house” is shaped by the features and experiences they associate with houses.

Doesn’t this support the idea that perception is selective and depends on the observer? The perceptual system stays connected to the world, but the organism does not need to take in or represent the whole scene at once. It can focus on information that matters for its current actions or needs—what is biologically relevant. Processing everything is unnecessary usage of resources.

I’m confused about what the author means by an “alternative conception of the locus of cognition.” I understand that standard cognitive science treats cognition as processes happening in the brain, but I’m not sure how an organism’s actions can extract invariant features from its environment without the brain doing the processing. Does this relate to how cognition is "offloaded"?

An observer does not simply receive a visual scene; they move their eyes and body to explore it. Those movements change the available visual information and help the observer detect what is stable in the environment. So Gibson reframes the binding problem: instead of asking only how the brain combines separate visual features, we can ask how the brain, body, and environment work together to make perception coherent.

Gibson’s answer differs from the idea that the brain receives separate bits of information and then assembles them. Vision depends on the whole perceptual system: the eyes move, the head and body shift, and information from vision interacts with information about balance and bodily movement. These actions reveal how the observer is positioned in the environment and how the environment changes as the observer moves. In this view, the brain helps the perceptual system pick up and use information; it does not create a complete picture of the world from the retinal image alone.

This conclusion strongly argues that cognition is only localized in the brain, but isn't energetically favorable for humans to extend their cognition by interacting with the environment? Offloading is a consequence of evolved human traits like memory limits, metacognition, etc.

How can researchers assess the reliability of the data collected through this method? Even while thinking aloud, participants still have to decide what to put into words. Some parts of their thought process may be instinctive, so they may not notice them or think they are worth mentioning. So, how accurately can this capture a person's overall thought process to solve?

To understand how symbol representation works, you need to know the rules used to operate symbols to transform input into output. But how does one replicate this system for higher order tasks? How can you generalize those rules to a larger group when it could differ based on geographical/cultural exposures?