Listen to this article
In recent days, I have been obsessively staring at telephone numbers. That is partly because I have just moved house and am flicking through my contacts list to send out change-of-address notes. But there is a second reason too: I have just stumbled on a fascinating little paper written by a Princeton cognitive psychologist called George Miller on the topic of “chunking”. And while this piece of research is half a century old, it has a curious relevance today – particularly in relation to those telephone numbers which are now so unthinkingly woven into the fabric of our 21st-century lives.
If Miller is correct, whenever we recite those digits, we unconsciously reveal the degree to which we are hard-wired to sort information into mental boxes. And that trend has important implications – even (or especially) though most of us never give a moment’s thought to the shape of those numbers.
The issue revolves around memory. Back in the early 1950s, Miller, like many psychologists and neuroscientists, was fascinated with the question of how brains retain information. Until that point, many scientists assumed that memory varied according to innate ability. However, Miller believed there was a more fundamental pattern. His research suggested that most people had a limit to how many pieces of data, such as numbers or letters, they could memorise when presented with a list. This usually ranged between five or nine data points but the average was “the magic number seven”.
There was a crucial caveat: if people learnt to group data they saw or heard into manageable chunks – or mental clusters of information – they could remember more, since each chunk became a data point in its own right. This worked particularly well when the clusters were associated with established ideas, or pre-existing mental “labels”. However, even nameless bundling helped too. “A man just beginning to learn radiotelegraphic code hears each dit and dah as a separate chunk,” Miller wrote, after conducting experiments among radio operators. “[But] soon he is able to organise these sounds into letters and then he can deal with the letters as chunks . . . [then] as words, which are still larger chunks, and [then] he begins to hear whole phrases.” Better still, over time the amount of data stored in each chunk expanded. “It is a little dramatic to watch a person get 40 binary digits in a row and then repeat them back without error. . . . [However,] recoding is an extremely powerful weapon for increasing the amount of information that we can deal with.”
Unsurprisingly, this chunking idea has spawned a plethora of academic debate in the past five decades. Some psychologists think the more natural number for chunking is four, not seven; others insist that chunking explains only a small part of how our memories work. But either way, Miller’s point – and legacy – live on in our numbers.
Most notably, if you ask an American to recite a phone number today, he or she will usually do it in three chunks, since that is how convention presents it. This is partly due to regional codes but it is also the pattern that seems to best suit our habits (and as companies such as AT&T were influenced by Miller, this is no accident). In London, people also quote numbers in three chunks. Elsewhere, there are different patterns, such as two chunks of five digits (in rural Britain) or several pairs (in France). Regardless, the key point is this: almost no one remembers numbers in a single string of unbroken digits. This feels unnatural.
Is this a good thing? In some senses, yes – chunking means we can remember more numbers, and much else. But I cannot help but wonder if there is not a darker side to this instinct too. By keeping things in mental boxes, our brains can also become dangerously rigid. To understand this, try quoting a telephone number you know off by heart in different chunks (say, a three-chunk New York number as two chunks of five digits). It is surprisingly hard.
While that rigidity – or habit – might seem just a curiosity in relation to phone numbers, it also extends to other mental processes. We are trained to keep all manner of data in mental boxes, in a way that often makes us prone to dangerous blind spots or prevents us from embracing lateral thought. Perhaps that is inevitable: we need order to live and survive. But the next time you pick up a phone, it is worth reflecting for a second on the way that we process and categorise data into chunks. If for no other reason than because “thinking out of the box”, as consultants say, can only occur when we remember the degree to which our mental processes are shaped by mental boxes, or chunks – especially when we do not notice those patterns at all.
Letter in response to this report:
Get alerts on FT Magazine when a new story is published