The Shallows: Chapter 9 Summary & Analysis

To begin a conversation about memory, Carr returns to Plato’s Phaedrus. In the Phaedrus, ancient orator Socrates warns that writing might cause people to be dependent on books and weaken their memories. Socrates’ fear, as Italian novelist Umberto Eco points out, was a natural and ancient one, “an eternal fear: the fear that new technological achievement could abolish or destroy something that we consider precious, fruitful, something that represents for us a value in itself, and a deeply spiritual one.” Socrates may have been right in part. However, Carr argues that books have historically freed people to chart an individual path of learning, and that books are responsible for man’s heightened focus on individuality.

What’s more, reading has proven to actually improve rather than deaden memory. To display this point, Carr writes that the Dutch humanist Erasmus advised students to memorize notable passages from their books. For Erasmus, memorization was not a mechanical process but a way to synthesize or internalize knowledge that speaks to you. Memorization fell out of fashion with increasing technologies for knowledge storage, and in the age of the Internet, we have a seemingly endless external database. As NYT columnist David Brooks puts it, we have “outsourced” our memory, putting us in the strange position of having access to everything and knowing less than ever before.

Carr turns his attention to the process of how memories are made. He begins by recalling scientist Eric Kandel, who demonstrated in the 1970s with his Aplysia experiments that synapses are altered by experience. Kandel, looking past the simple withdrawal of the slug’s gill, wanted to investigate the underlying issue: How the brain was transforming experience into memories. In 1890, philosopher William James concluded there were two kinds of memories: primary memories, which we forget almost instantly, and secondary memories, which we can remember forever.

Studies on boxers who develop amnesia after blows to the head imply that even strong memories remain unstable after they are formed. Further research suggests that the brain requires a certain period of hours to “fix” a memory and transfer it from short-term to long-term. The process is delicate and any disruption can erase the memories forever. In fact, the storage of long-term memories, as proved by U Penn neurologist Louis Flexner, is biological, requiring the synthesis of new proteins, whereas the creation of short-term memories is not. Kandel’s continued research on the sea slug, in which he traced the neuronal signals, not only proved that repetition of an action encourages the consolidation of a short-term memory into a long-term memory, but also cast light on Flexner’s discovery. Kandel found that the creation of long-term memories stimulated growth of new synaptic terminals. In other words, the anatomy of the brain had to change in order to store the long-term memories, proving––as Kandel wrote in his 2006 memoir––that the anatomy of the brain is changed with learning.

Carr illuminates two other types of memories: implicit and explicit. Implicit memories are recalled automatically from the unconscious when doing a performative action like riding a bike. Explicit memories are recollections of facts and happenings in our past experienced in conscious working memory. Carr points out that the memories we are usually referring to when we talk about our memories—in this book and in general—are the explicit ones.

Carr notes that when storing explicit memories, or consolidating them into long-term memory, an ancient part of the brain called the hippocampus plays a pivotal role. In 1953 a man named Henry Molaison had part of his hippocampus removed to cure his epileptic seizure. Unfortunately, though his seizures stopped, Molaison was unable to remember many of his recent explicit memories and was no longer able to store new ones at all. His experience suggests that the hippocampus is the holding place for new memories. Once the memory is fully consolidated, it is sent to the cortex for secure storage––but the process can take years, explaining why so many of Molaison’s memories vanished.

Carr’s in-depth explanation of human memory consolidation serves to highlight the problems with an analogy that compares human memory to computer memory. Human memory, unlike computer memory, is alive––it is a biological process. Carr quotes Kobi Rosenblum, an extensive researcher on memory consolidation: “While an artificial brain absorbs information and immediately saves it in its memory, the human brain continues to process information long after it is received, and the quality of memories depends on how the information is processed.”

The idea of outsourcing memory, then, Carr suggests, is invalid because memories have a unique history that changes each time we recall them. If a memory is brought back into working memory, it turns back into a short-term memory and so gains a new context. Biological memory is in a perpetual state of renewal, where memories change each time they are moved from one place in the brain to another. In contrast, computer memory is comprised of fixed items that always stay exactly the same, no matter how many times you move them back and forth. Also unlike a computer, the human brain has no storage cap. Our cognitive powers aren’t constrained when we store new long-term memories and we can, conceivably, keep storing new ones forever. The idea, then, that online databases free our brains for intelligent thought by outsourcing memory is flawed because the two types of memory function so differently.

The Internet, in fact, places so much pressure on working memory that consolidation of long-term memories is obstructed. Consolidation of long-term memory depends on our level of attentiveness. As Kandel writes: “For a memory to persist, the incoming information must be thoroughly and deeply processed.” If the working memory is overloaded and we are unable to attend to it, that information is released in a matter of seconds––meaning the Internet, by overloading working memory, is not helpful to the consolidation of long-term memory.

Worse yet, due to neuroplasticity, the more we use the Internet the more we train the brain to process information quickly but without attentive focus. We become, in other words, very good at forgetting and worse and worse at locking information into our biological memories. This creates a feedback loop in which our trained-to-forget memories rely increasingly on the Net’s databases. Carr’s point is that the connections of the Internet are not remotely as rich and complex as our own synaptic connections. As Carr puts it: “When we outsource our memory to a machine, we also outsource a very important part of our intellect and even our identity.” What is at stake is the very nature of our identities. We risk becoming spread as thin as the Internet––becoming “pancake people”––without the complex internal architecture of personality taken as a given in the decades before we had such overwhelming access to knowledge.

Digression. Carr addresses the issue of how he was able to write this book at all in the age of distraction. Carr writes that at the beginning he struggled immensely, only able to write in spurts and constantly distracted by the Net. He made a drastic change in order to really get the work done and moved to an isolated house in the mountains of Colorado. There, Carr only had a slow DSL connection and no cell service. He canceled his social media connections and kept his email program turned off for the majority of the day. Dismantling his life on the Net caused definite withdrawal pains, but eventually Carr felt like his brain readjusted to literary thinking. He began to calm down and regained the ability to focus on his work.