Chapter 9: Active Verbal Memory: Encoding, Structure, and Recall

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to the Deep Dive.

Today we are opening up a true classic in the field.

We're essentially defined the field as we know it.

I mean, it's a foundational text.

It really is.

And we are zooming in on a topic that is, well, it's happening right now in your brain.

As you listen to this very sentence, we're talking about active verbal memory.

It's the engine of consciousness in so many ways.

You could call it the workspace where we process the world.

Exactly.

And just to set the stakes here, our mission today is to do a really comprehensive step -by -step walkthrough of this specific chapter.

We aren't just skimming the surface or, you know, pulling in some random Ted Talks.

No, this is about the source material.

We're going to deconstruct how your mind holds onto information exactly as the text presents it.

So whether you're a psychology student trying to figure out that this is between decay and interference for an exam or just someone wondering why you forget a phone number the second someone interrupts you, this is the Deep Dive for you.

And it is so vital to define exactly what we mean by active verbal memory right at the start, because it's not just memory in the sense of, you know, remembering your childhood home or your wedding day.

Right.

The text calls it immediate memory.

It does, but it makes a very specific and I think very important distinction.

It describes it as auditory synthesis extended over time.

Okay.

Auditory synthesis extended over time.

That sounds incredibly technical.

Let's unpack that.

Think about it practically.

Language happens in time.

It's linear, right?

Sure.

So if I speak a long, complex sentence to you, by the time I get to last word, the first word is it's gone.

Physically, the sound waves have just vanished into the air.

So if your brain operated purely on what the chapter calls echoic memory, just that passive echo of sound like a microphone recording on a tape, you'd lose the beginning of my sentence before I even finish it.

Exactly.

You wouldn't be able to comprehend complex language at all.

You need a workspace.

You need a medium that actively holds onto that information, synthesizes it, and well, keeps it alive just long enough for you to extract the meaning.

And that's active verbal memory.

That's it.

It's the bridge between just hearing a noise and actually understanding a thought.

And the chapter makes this fascinating point that this isn't just for listening.

It shapes how we read too.

Oh, absolutely.

Even when the input is visual, like you're reading a textbook or a novel, we are so often translating that visual code into a verbal description just to hold on to it.

We're saying the words in our head.

We're saying the words in our head to keep them in the workspace.

It's all about that internal voice.

So give us the roadmap.

This is a dense chapter.

There's a lot of heavy lifting here.

What are the big landmarks we are going to hit in this deep dive?

Well, we've got a lot of ground to cover.

We're going to start with the memory span and the famous magic number seven, but we're going to look at it through the lens of information theory, which is where you see that the brain is actually, it's kind of weird mathematically.

And from there, we'll talk about how we hack that limit with recoding and chunking.

Then we get into the nuts and bolts, like what does a thought sound like?

We'll look at some really clever evidence that the storage medium is effectively auditory.

Then we'll watch a kind of championship fight between two theories of organization, slot theory versus association theory.

And I know you're excited about the rhythm section.

I am.

It might just be the missing link in how we organize our minds.

Then we'll get into why we forget.

Is it just time passing or is it noise?

And finally, we will wrap up with the duplex theory.

Is this short -term workspace a totally different system from your long -term memory, or are they connected?

Fantastic.

Let's dive right in then.

Section one, the capacity and the code.

Let's do it.

So the starting point for all of this is the span of immediate memory.

This is one of the oldest measurements in experimental psychology.

If I read you a list of random digits and you have to repeat them back, the average normal adult can handle about seven.

Give or take two, it's remarkably consistent across people.

The text actually quotes David Wetzler, the guy who created the IQ tests we still use today.

Oh, right.

He said that unless there's an organic disease, an adult who can't retain five digits forward is, in nine cases out of ten, feeble -minded.

Wow.

That is some harsh terminology from the past.

It is, for sure.

But it shows you just how standard this measure is.

It's a baseline for cognitive health.

If you can't hold five things in your head, the thinking was, the system is broken somewhere.

Okay.

But here's where the text throws a wrench in the works.

In the mid -20th century, psychology was, well, was kind of obsessed with information theory.

Totally of so.

We've touched on this in previous Deep Dives.

This is the whole idea of measuring information in bits, like a computer.

Correct.

A bit is a binary choice.

It's a yes or no, a zero or a one.

It's a unit of uncertainty reduction.

Now, from a computer science perspective, different types of data contain different amounts of bits.

Right.

So let's look at a single decimal digit, you know, zero through nine.

That's one option out of ten.

That represents a certain amount of information.

Yep.

Mathematically, it's about 3 .32 bits.

Now, compare that to a letter of the alphabet.

That's one out of 26.

That contains more information, more bits, because there are more possibilities to distinguish between.

And a word, a monosyllabic word chosen from thousands in the dictionary.

I mean, that represents a massive amount of bits compared to a single digit.

Okay.

So here's the hypothesis then.

If the human brain is like a hard drive with a fixed bandwidth, a fixed number of bits it can hold,

we should see a huge difference in how many items we can remember based on what they are.

Exactly.

If you have a strict bit limit, you should be able to remember, say, dozens of digits, maybe a dozen letters, but only two or three words.

The span should fluctuate wildly depending on the information density of the item.

But that's not what happens at all.

No.

It's completely flat.

The memory span is roughly the same.

About seven items, whether those items are digits or letters or short words, the span is invariant to the bit count.

So the brain effectively just ignores the bit value.

It ignores the mathematical information content completely.

And this is the key takeaway.

The brain doesn't measure capacity in bits.

It measures capacity in chunks.

The chunk.

This is George Miller's famous contribution, the magic number seven.

It is.

A chunk is a cognitive unit that's created by you, the subject.

The limit of immediate memory isn't bits.

It's five to nine chunks.

That's Miller's magic number seven plus or minus two.

It doesn't matter how much information is inside the chunk.

It only matters that it counts as one thing to your brain.

The text gives a really great example of this with sentences.

If I give you 20 random words, you probably can't repeat them back.

That's 20 chunks.

It's just too many.

Way too many.

But if those 20 words form a coherent sentence,

the tall man walked down the street to buy a newspaper,

suddenly you can repeat it perfectly.

Why?

What changed?

Because you aren't remembering 20 separate units anymore.

Your brain groups them.

The tall man becomes one chunk.

Walk down is another.

The street is another.

You're basically compressing the file.

You turn 20 individual words into four or five meaningful phrases.

So you're still obeying the limit of seven.

You're just packing more data into each slot, so to speak.

Exactly.

You are enriching the content of the chunk, but the number of chunks remains the same.

Now, there's a skeptical view here.

You could argue, well, maybe the brain just gets better at hearing sentences because we hear them all day long.

But the chapter details this fascinating case study that proves this is an active, deliberate process.

The case of S .L.

Smith.

Ah, yes.

This is one of my favorite experiments in the history of psychology because it shows just how malleable the mind is when you apply an actual strategy.

So S .L.

Smith was a graduate student who decided to test his own memory span for binary digits.

Zeros and ones.

Just a stream of zero, one, zero, zero, zero, one, zero, one.

That sounds maddening.

I can't even imagine.

So what was his baseline?

His natural span for these binary digits was about 12.

He could hear 12 zeros and ones and repeat them back perfectly.

Which is already impressive, honestly.

That's higher than the standard seven.

It is, though binary digits are very simple and repetitive, so it makes some sense.

But Smith wasn't satisfied.

He wanted to see if he could hack the system using this chunking theory.

So he learned to recode.

Okay, explain recoding.

This feels like a crucial concept in the chapter.

It is.

He basically memorized a translation system.

He learned to group the binary digits he was hearing into triplets groups of three and convert them into octal digits.

Octal digits.

What are those?

It's a base eight number system.

It sounds complicated, but it's really just a mental mapping table.

For example, if you heard the triplet zero, zero, one, he would translate that in his head to the number one.

If you heard zero, one, zero, he'd translate that to two.

If you heard one, one, one, that becomes seven.

Okay, I see.

So he's listening to this rapid stream of binary and he is actively doing math in his head to convert them into new numbers and then memorizing those new numbers.

Yes.

He is reformatting the data stream in real time.

He isn't just passively listening.

He's transforming it.

So what happened to his memory span?

It just skyrocketed.

His binary span went from 12 to nearly 36 digits.

36 digits.

That is superhuman.

If you just saw that number in a vacuum, you'd think he had a photographic memory or some kind of genetic advantage.

It looks superhuman on the surface, but the active verbal memory analysis tells us exactly what was happening.

He wasn't actually remembering 36 separate items.

Okay.

Let's do the math here.

36 binary digits.

He breaks them into groups of three.

That's 12 groups.

Correct.

So he converts those 12 groups into 12 octal numbers.

He remembers the 12 numbers.

Then when he has to recall them, he just translates them back into binary.

So he was still remembering roughly 12 chunks.

He didn't change the fundamental capacity of his brain.

He just changed the density of the package.

Jerome Brunner put it beautifully.

He said, Smith made the chunks filled with purer gold.

This is the definitive proof that active verbal memory involves reformatting the input.

You don't just echo what you hear.

You actively change it to make it stick.

But this recoding process, this translation, it takes effort, right?

It takes cognitive load.

And crucially, it must take time.

It absolutely does.

You are performing a mental operation, not just passively recording.

Which leads us to the next big question.

The chapter tackles the rate of presentation.

If I have to do all this mental gymnastics to recode information, does it matter how fast you say the list to me?

This is where we see a fascinating conflict between two systems in the brain.

It's a trade -off.

On one hand, you have that echoic memory we mentioned, that passive sound echo that fades very, very fast.

So for the purely echoic memory, speed is actually good.

You want to get the whole list in your ear before the beginning fades away.

Right.

If I speak too slowly, the start of the sentence has evaporated from the echo before I even get to the end.

But for active memory, the recoding part speed is bad.

Because I don't have enough time to do my translation.

If I'm S .L.

Smith, I need that split second to turn 001 into the number one.

Exactly.

And there's a study by Intima, Wozencraft, and Clem that illustrates this perfectly.

They used a computer to speak digits at an incredible rate, 10 digits per second.

10 per second.

The text says it sounds like ripping cloth.

It's just a blur sound.

Burp.

But each digit was, technically, intelligible.

Now, at that speed, there is zero time for recoding or grouping.

You are relying entirely on that passive echo.

And the result?

What happened to recall?

It dropped drastically.

It fell off a cliff.

Subjects could only get about three or four items right.

They were just completely overwhelmed.

But when they slowed it down to two digits per second, recall went right back up to normal near six or seven.

So slow is smooth and smooth is fast, in a way.

In this context, yes.

Slow presentation is almost always better because it buys you time.

Time to organize, time to group, time to recode.

The conclusion is that while that echo helps, active verbal memory relies heavily on this reorganization process.

If you go too fast, you break the machine.

You're essentially flooding the engine before it can process anything.

Precisely.

This brings us really nicely to section two.

We've talked about capacity and the process of chunking.

Now we need to talk about the medium itself.

What does a thought sound like?

This is one of those deep, almost philosophical questions that cognitive psychology managed to answer empirically.

When we hold a phone number in our head, is it a visual image of the numbers?

Is it some abstract concept or is it a sound?

The chapter is pretty firm on this.

It says the storage medium is essentially auditory and linguistic.

Yes.

This is true even if the input is visual.

If I show you a list of letters on a screen, you don't store the picture of the letter.

You store the name of the letter.

You store the sound.

The evidence for this is just so clutter.

It comes from Conrad back in 1964.

It's all about acoustic confusions.

This is a classic, classic experiment.

Conrad showed subjects letters visually.

They saw them.

They didn't hear them.

Then he analyzed the mistakes they made.

The logic here is just beautiful.

The type of mistake you make reveals how you are storing the data.

Right.

If I'm remembering the visual shape of the letter and I see a B, I might mistake it for an R because they look kind of similar.

They both have that vertical line in the loops.

Exactly.

A visual storage system should produce visual errors.

E for F, O for Q.

But that's not what happened.

Subjects almost never mistook B for R.

Instead, they mistook B for P or D for T.

Because they rhyme?

Because they sound alike.

The acoustic confusion tells us that the visual input was immediately translated, recoded into an auditory representation.

The errors were hearing errors, not seeing errors, even though the input was purely visual.

That is wild.

It explains why when I'm reading a really tough sentence, I often have to say it in my head.

I'm converting it to the file format that my brain's RAM understands.

And Wickelgren took this even further.

He didn't just look at errors.

He looked at difficulty.

He created lists of letters that all shared a vowel sound.

For example, F, L, M, N, S, X.

E, F, L, M, S, S.

Okay, yeah, they all start with that short A sound.

Right.

And he found that lists like that are much, much harder to recall than lists with distinct sounds like, say, J, K, R, Z.

He calls this phonemic similarity.

Because the internal representations are all jumbled up and overlapping.

They get muddy.

And the theory here, the really big idea, is that immediate memory uses the same mechanism as speech perception.

It's called analysis by synthesis.

Okay, break that down for me.

To understand speech, the theory goes, we partially mimic the motor commands that would be required to produce that speech.

And to remember a list, we say the items to ourselves.

We re -synthesize the speech sounds internally.

So if the sounds are too similar, like that list of F, L, M, the internal synthesis gets jam -packed and errors happen.

So thinking of a word is, in a very real neurological sense, speaking it silently.

It's inner speech.

Absolutely, that's the medium.

Okay, so we know we have a limit, which is chunks.

And we know the format is auditory, it sounds.

Now, section three asks the next logical question.

How is it organized?

We have these seven chunks floating in our head.

How do we keep them in the right order?

This is the championship fight I mentioned earlier.

We have two competing theories trying to explain how we keep the items in serial order.

In the blue corner, we have slot theory.

In the red corner, association theory.

All right, let's break down slot theory first.

It seems the most intuitive on the surface.

It is.

Just imagine a row of mailboxes or bins inside your head.

Let's say you have seven of them.

That's your memory span.

Okay, I've got my seven bins lined up.

When you hear a list, you put the first item in bin one, the second item in bin two, the third in bin three, and so on.

Simple enough.

It has some strong arguments going for it.

For one, it explains why we know where an item was in the list.

If I ask what was the third number, you just look in bin three.

It also explains transposition errors.

That's when you swap items, right?

Like if the list was six, nine, four, and you say six, four, nine.

Exactly.

In slot theory, you could argue the item just hopped into the neighboring bin.

It fell into the wrong slot.

But the text seems to think slot theory is a bit too passive, a bit too rigid.

It is.

It's too rigid.

It doesn't account for the fact that sentences vary wildly in length.

If you have exactly seven fixed slots, how do you remember a 10 -word sentence?

Do you just sprout new slots?

And why do we group things?

A slot system shouldn't care about rhythm or phrasing, but we humans definitely do.

Okay, so what's the contender?

Association theory.

Association theory says there are no bins.

Instead, memory is a chain.

Item one is linked to item two.

Item two is linked to item three.

Like a daisy chain.

Or dominoes falling.

Exactly.

Hearing the number six triggers the association with nine, which triggers four.

This makes a lot of sense because it connects short -term memory to long -term memory.

We know long -term memory works heavily by association.

That's its biggest strength.

But association theory has a, well, pretty fatal flaw.

It's called the end -of -the -list problem.

Explain that.

Well, if item one triggers item two and two triggers three, how do you know when to stop?

There is no stop stimulus.

If I say a list of numbers and then I just stop talking, how does your memory chain know that the last number is the last number?

The chain just runs out of links.

But in a purely associative theory, that last item might be associated with something else from your past.

If the last number is five, why doesn't your brain just keep going to six, seven, eight?

What puts the brakes on the process?

I see.

And it also struggles with repeated items, right?

Yes.

This is Wickelgren's evidence on something called associative intrusions.

Imagine a list like this.

Nine, two, nine, five.

So the number nine appears twice.

In the chain, the first nine is linked to two.

The second nine is linked to five.

Wickelgren found that people make a very specific type of mistake here.

They might say nine, five when they should have said nine, two.

Because the number nine is now associated with two different followers, the brain gets confused and picks the wrong path down the chain.

So they're coming to a fork in the road and taking the wrong turn.

Precisely.

But here's the problem.

We don't make that mistake every time.

We usually know that the first nine goes with two and the second nine goes with five.

Pure association theory has a very, very hard time explaining how we keep those duplicate items straight without some sense of position.

So slot theory is too rigid, but association theory gets lost to the crossroads and can't explain position.

Which brings us to the verdict in the text.

Neither one is fully satisfactory on its own.

We need a third way.

And that is section four, rhythm and structure.

This is the alternative view proposed in the chapter.

This is, for my money, the most exciting part of the chapter.

It proposes that the memory span isn't a row of boxes and it isn't a simple chain.

It is a rhythmic structure.

The text notes that subjects almost always group digits.

If I give you a 61935827, you don't say 61935827.

You say 61935827.

You impose a rhythm on it, da da da, da da da da.

You create a pattern.

Why does this work?

Why does adding a rhythm, which is technically more information to deal with, make it easier to remember?

Because rhythm creates a gestalt.

It creates a whole organized structure.

The text references Carl Lashley, a famous sonorous psychologist, and he pointed out that rapid skilled responses, like playing the piano or even just speaking, are way too fast to be chained one by one.

You don't think C then E then G?

You think C major chord?

You structure the entire sequence in advance.

The chapter suggests that rhythm creates pseudoslots.

If you have a rhythmic pattern in your head, da da da, you have created a structure with three empty spaces.

You can then hang the digits on that structure.

So the rhythm provides the slots that slot theory wanted, but they are flexible and they are created by the user, not built in.

Exactly.

This explains why you know where an item is.

You can say, oh, it was on the first beat of the second triplet.

It also explains why backward recall is so ridiculously hard.

Oh, that makes so much sense.

If I learn a song, I can't easily sing it backward.

I have to completely break the rhythm, the entire musical structure.

Precisely.

You have to destroy the structure you created in order to retrieve the items in reverse.

It also explains the running memory span problem.

That's where I read you a list, but I don't tell you when it's going to end.

Right.

And if you don't know the length, your recall is terrible.

Why?

Because you don't know what rhythmic structure to build.

Do I build groups of three, groups of four?

You can't build a scaffold if you don't know the shape of the building.

There is a really stark example of how fragile this rhythm is.

Conrad's knot experiment.

This is such a simple but devastatingly effective experiment.

So Conrad had subjects listen to a list of eight digits, but he asked them to do one simple thing.

Just say the word knot, which is just the British word for zero, between hearing the list and recalling it.

Just say one word, knot.

That's it.

One single word.

What happened?

Accuracy dropped from 73 % to 38%.

That is a massive crash just for saying one word.

Think about it rhythmically.

You've got this delicate, self -imposed rhythmic structure holding the digits.

Da da da da.

Then you force yourself to insert an extra offbeat verbal impulse, knot.

It shatters the rhythm.

It breaks the very structure that your active memory relies on.

It's like someone clapping offbeat while you're trying to dance.

It just messes everything up.

That is a perfect analogy for it.

This leads us naturally to section five then.

Forgetting.

Why do we lose these numbers?

Is it just time passing or is it this kind of active interference?

But the great debate.

Decay versus interference.

Let's visualize the key piece of evidence here.

I want you, the listener, to picture a graph.

It's figure 41 in the text.

It's the famous Peterson and Peterson curve.

So on the vertical axis, you have percent correct recall from zero to 100.

On the horizontal axis, you have time in seconds from zero to 18.

And the procedure is simple.

You hear three consonants, something like chj.

Then immediately, you have to start counting backward by threes.

300, 297, 294.

And that counting is crucial.

It's there to stop you from rehearsing.

It occupies your inner speech, your articulatory loop.

And what does the line on the graph do?

It just drops like a stone.

At zero seconds, you're near 100 % recall.

After just 18 seconds of counting, recall is down to barely 10%.

It looks exactly like the memory just evaporated into thin air.

And that was the initial interpretation.

This was held up as the classic evidence for decay.

The idea was that the memory trace is like a footprint in the sand.

Without rehearsal to stomp it back in, the wind of time just blows it away.

But the chapter argues pretty strongly against simple decay.

It does.

It argues for interference.

The argument is it's not just that time passed.

It's what you did during that time.

The counting backward.

Right.

And Wickelgren, who, as you can see, is all over this chapter, showed that if the intervening material sounds similar to the target material,

the forgetting is much, much worse.

This is acoustic interference.

So if I'm trying to remember letters, and you make me count numbers, that's bad.

But if I'm trying to remember letters, and you make me read other letters that sound similar, that's disastrous.

Yes.

If it was just time causing decay, it shouldn't matter what the distraction is.

But it matters a lot.

Similar sounds cause more confusion, more interference.

And there's also the concept of proactive inhibition.

That's when previous lists confuse the current list, right?

Exactly.

If you do 10 trials of this Peterson and Peterson experiment, by the time you get to the 10th trial, you're confusing the current letters with the letters from trial 1 and 2 and 7.

The arbed memories are actively interfering with the new ones.

So what is the synthesis here?

Is it decay or is it interference?

The expert in the text proposes a really compelling compromise.

Remember the difference we talked about between echoic memory, the passive echo, and active memory, the recoded structure?

Yes.

The conclusion is this.

Echo of memory likely decays.

The raw, unprocessed sound just fades away.

But active verbal memory suffers from interference.

Ah.

So once you've built that rhythmic structure, that active thought, it doesn't just fade.

It gets smashed by other similar thoughts.

Exactly.

And this redefines what rehearsal really is.

We usually think of rehearsal as just refreshing the paint on the wall to keep it from fading.

But this view says rehearsal is the active process of resynthesizing the material into a new structure over and over.

So you aren't just preserving the old memory.

You are literally rebuilding it from scratch again and again.

Yes.

And if you are interrupted by counting backward or saying the word not, you can't perform that active rebuilding.

And so the structure collapses.

That is a really powerful distinction.

It changes how I think about study.

Honestly, it's not about just holding information.

It's about constantly rebuilding it.

Active processing.

That's the key to everything.

OK.

We have one final section to unpack.

Section six, two kinds of memory or one.

The duplexity theory.

Is this short -term active memory a completely different system from long -term memory?

Or is it all just one big spectrum of activation?

Broadbent was a major proponent of the two systems view.

He proposed an S system, a short -term buffer, and a P system, the perceptual system for long -term storage.

And he had a really cool experiment to support this called the split -span.

Imagine wearing headphones.

In your left ear, you hear three digits, say six, four, nine.

At the exact same time, in your right ear, you hear three different digits, like two, eight, three.

So I'm getting blasted with numbers from both sides simultaneously.

That sounds incredibly stressful.

It's tricky.

Now, if I ask you to tell me what you heard, you will almost always report them year by year.

You'll say six, four, nine.

And then you'll say two, eight, three.

You won't mix them.

You won't go back and forth and say six, two, four, eight, nine, three.

People find that incredibly difficult.

And Broadbent argued, this is proof there's a buffer.

You process the left ear immediately, while the right ear's information sits in a temporary holding tank, the S -system waiting its turn to be processed.

That suggests a very distinct short -term mechanism.

It holds the line while the main processor is busy.

But then we have Hebb, and Hebb's work kind of bridges the gap.

Hebb's recurring digits experiment is so fascinating because it gets at the subconscious.

He gave subjects many, many strings of nine digits to recall over and over again.

Just random strings of nine digits?

Well, they thought they were random.

But unknown to them, every third string was identical.

A secret code.

A secret pattern embedded in the noise.

And here's the result.

Subjects gradually got better and better at recalling that repeating string.

By the end of the session, they were nailing it.

Did they know it was repeating?

Many of them didn't.

Even if they had no conscious realization that a string was repeating, their brain learned it.

So something that started in short -term memory left a permanent trace in long -term memory, just through sheer repetition.

Exactly.

This implies that the two systems aren't completely sealed off from each other.

Short -term repetitions can carve a groove into long -term memory.

It suggests a deep connection.

Or perhaps that short -term memory is just the active portion, the activation of long -term pathways.

So we've covered a massive amount of ground here.

Let's try to bring it all back home for the listener.

Yeah.

What are the big essential takeaways from chapter nine?

Okay.

If you walk away with four things, make them these.

One, active verbal memory is fundamentally auditory and linguistic.

It's inner speech.

Two,

your capacity is determined by chunks, that magic number seven, not by bits of information.

Three,

recoding and rhythm are the active tools we use to hack that capacity.

We group things and impose a structure on them to make them stick.

And four.

And four, forgetting, at least in this system, is largely due to interference, especially acoustic interference, which breaks down that rhythmic structure you've built.

Yeah.

It really highlights how incredibly active and structured our minds are.

We aren't passive sponges just soaking up information.

We are builders.

We take sound, we break it into chunks, we build a rhythmic scaffold, and we constantly have to rebuild it to keep it from collapsing.

And that leads to a final provocative thought that the chapter hints at.

We talked about how important rhythm is for memory.

That da -da -da, da -da -da pattern.

Right.

Well, just consider language itself.

We speak in phrases, we have prosody, we have cadence and rhythm in our speech.

If active verbal memory relies on rhythm and structure,

well, it suggests that spoken language itself might just be a highly complex pre -structured rhythmic memory exercise.

So we speak the way we do, because that's the only way we can remember what we're saying while we're in the middle of saying it.

Exactly.

The structure of language mirrors the structure of memory, perhaps because it has to.

That is something to mull over the next time you find yourself pausing in the middle of a sentence.

Thank you so much for joining us on this deep dive into active verbal memory.

It's always a pleasure.

And to our listeners, keep those neurons firing, keep building those chunks, and we'll catch you on the next deep dive.