Does memorisation get in the way of learning? – Part 3

Yesterday I said that these were the main points in Ben Orlin’s article expressing concern that ‘memorisation might get in the way of learning’:

    • Some things are worth memorising, but others are not.
    • Knowledge matters – “even a head full of memorized facts is better than an empty one.”
    • Raw rehearsal is a memorisation technique, but not a very effective one.
    • Mnemonics are a better memorisation technique, but they still promote the memorisation of meaningless facts.
    • ‘Repeated use’ is a better route to memorisation, and one that he advocates.
    • ‘Building on already known facts’ is a second route to memorisation that he strongly advocates.

I started by arguing that memorisation is learning.  Remember the quote from Kirschner et al.’s paper: “If nothing has changed in long-term memory, nothing has been learned.”

Given then that some teachers express concerns regarding memorisation that are completely valid, I next looked at the first two points that Orlin makes, and attempted to explore them further.  I summarised by introducing a new category of knowledge that could be confused with ‘rote knowledge’: ‘inflexible knowledge,’ and echoing Willingham’s statement that it is the necessary first step on the path to what we usually think of as ‘understanding.’

Today I would like to look at the next two points, criticisms of what Orlin calls raw rehearsal and mnemonics.

Raw Rehearsal

This is effectively what most of us would think of as ‘cramming.’  If you wanted to memorise the first 50 names of dinosaur species alphabetically, you could do it by staring at the list, reading it, re-reading it, covering parts and trying to recall them etc. until you get them all.

Problems with this?  Having memorised a speech in this fashion for ninth grade English, Orlin makes this point:

“The process was slow, dull, and stilted. I forgot the speech within weeks.”

My concern here is that Orlin is right, or at least appears to be right.  ‘It’s boring; we forget it straight away.’  These things can be true.  The issue I take is that they don’t have to be true.  Let me offer a counter example.


Rehearsal, a slow, dull and stilted process

(okay maybe I’m being a little facetious)

After my first year in university, I did a summer job selling frozen food door to door.  We were asked to memorise a script that we would use when first speaking to people at the door.  We were asked not to deviate from the script wherever possible; the words had been carefully chosen over a long period of time and iteration in order to have maximum positive impact on a stranger we were meeting for the first time.  We memorised and practised it, having some fun doing so.  Ten years later I can still recall it perfectly.

Months ago I mentioned actors using this exact technique of ‘rehearsal’ to memorise vast quantities of text.  More recently I talked about memorising poems; while for some of them I used an app that employs a few clever tricks, in other cases all I had was the poem on the screen, and how else could I memorise it but the technique Orlin attacks?  It can be very effective.

Orlin is right, it can be boring, bland, and can lead to the kind of forgetfulness we’ve all experienced after walking out of an exam.  But he’s missing a trick; it can also be a useful tool.  What is he missing?  Why did he forget his speech, while I still recall my script after so long?  Why do I recall my poems when he forgot his speech by the same method, and why did I have fun learning them in the first place (note: it doesn’t feel wrong for me to say ‘learning’ there, does it?  Memorisation is learning.)

The problem isn’t with the technique; the problem is with the strategy.  At least one piece of strategy is missing from the scenario Orlin described: distributed practice.

Orlin’s is the classic case of cramming.  He built the retrieval strength of the memory, but not its storage strength.  In this instance, he will have been able to recall the memory for the next day, but then quickly stopped thinking about it, and without any storage strength for the memory, rapidly lost his ability to recall it.

Storage + Retrieval

A slide from my talk on memory at TLT13

Storage strength can only be built over time, lots of it.  Distributed practice is a way of doing this.  If you only have four hours to practice something, or learn or memorise something, don’t do it all in one four-hour block – spend half an hour on it every day, or every other day, across 8-16 days.  This helps to build the storage strength of the memory.  What did Orlin do?

“When I had to memorize a speech for ninth-grade English, I huddled in the school library for 90 minutes, whispering the words to myself again and again, until they settled into my memory.”

In this example at least, he used ‘massed practice,’ cramming all his practice into one 90 minute block.  It gives the appearance of learning; as he says, the words ‘settled into his memory,’ but in fact all it does is build short-term retrieval strength.  It was perfectly predictable that he would quickly forget it all.

What was different in my examples?  Well I spent the next 9 weeks saying that script between a dozen and 80 times each day.  With the poems, every now and then I recite them to myself in my head; at first quite regularly, and later every few days, or every few weeks.

* ‘Raw rehearsal,’ as Orlin calls it, is not the enemy.  Massed practice is.

*  Distributed practice builds storage strength.


Distributed practice builds storage strength


Orlin describes Mnemonics as ‘artificial tricks.’  He argues that they are more effective than ‘raw rehearsal,’ but has this problem with them:

“They still bypass real conceptual learning. Memorizing a list of prepositions isn’t half as useful as knowing what role a preposition plays in the language.”

This deeply, deeply troubles me.

Mnemonics are an exceptionally powerful tool, and for them to come under attack is very concerning, especially at a time when teachers aren’t already being explicitly encouraged to use them, or taught how to do so effectively.  Given what is written about mnemonics, and given their demonstrable efficacy in a range of contexts, I can only interpret Orlin’s writing here as that of a layperson with some experience and an opinion, and he has it dead wrong.

A few examples:

Michel Thomas makes extremely heavy use of very effective mnemonics in his language courses.  Having heard once that ‘acheter’ (pronounced ‘ash-e-tay’) means ‘to buy,’ because it’s like going into a shop and saying ‘I’d like to buy that ashtray,’ I never forgot it.  While I was still learning the basics of French, that mnemonic helped me recall the word when I needed it.  Now?  Well by now my mind has fully formed the connection between the word ‘acheter’ and the concept of ‘buying,’ so the mnemonic is no longer needed.


“I’d like to buy that ashtray”  –  ‘to buy’ -> ‘acheter’

Harriet Ball, mentor to the KIPP school founders, similarly used many mnemonic devices, referring to them as ‘disposable crutches’ – you use them until you don’t need them anymore.  Levin and Feinberg refer to their sense of awe at how she could help young minds remember almost anything in the book Word Hard, Be Nice.  Here’s a great example of a disposable crutch:


I’m sure if you had any concern over showing crying elephants to young children, the mnemonic could be adapted.  The point is that the last thing we want is anyone using a mnemonic every time they spell the word ‘because’!  However, the only other way to learn that correct spelling is through persistent, deliberate and preferably distributed, practice (‘rehearsal’), and if when you first start that practice, you’re not actually sure of the correct spelling… well then you won’t be able to practise correctly!  So, we need something to rely on, until the correct spelling finally becomes a part of long-term memory without the need for the mnemonic anymore.

Daniel Willingham – who arguably seems to understand the inner workings of a learner’s mind more than most! – offers a whole range of mnemonic devices to help a novice learn (memorise) a new concept, a new idea, a new fact.

For me though, the real killer blow comes from Hattie and Yates’ latest book, Visible Learning and the Science of How we Learn:

                 “So the research question became, would students who are taught by High Mnemonic Teachers display any measurable learning advantages across the course of their year?

                The answer, as emerging straight from research findings, appears a clear ‘yes’.”

When Hattie and Yates, can make such an unequivocal statement, in addition to the other examples from experience above – when they seem to be saying in no uncertain terms ‘teachers who make heavy use of mnemonics will be better teachers than those who do not’ – it greatly troubles me that someone would advocate that teachers not use them, especially when that person is being promoted by someone with Meyer’s level of influence.


Mnemonics help us to recall things, until we don’t need them anymore

More worryingly, as before, Orlin’s argument approaches the truth.  This is to be expected – he is an intelligent person (graduated from Yale) with five years’ teaching experience; he’s not likely to say something without reason!  But, because what he says looks so close to the truth, it can easily convince others to think that it is the truth, and the whole truth.  Let’s look at Orlin’s argument again:

Orlin argues that a mnemonic is taking the place of meaningful understanding.  While he’s right to suggest that this can be the case, I strongly believe yet again that he is wrong to argue that this necessarily is the case.  For example, I like to think I now have a pretty good ‘understanding’ of trigonometry, but I still make heavy use of the SohCahToa mnemonic.  The last thing I want to be doing each time I simply want to decide which ratio to use, is to be wasting time deriving it from first principles on each occasion!  In his example of the prepositions, well yes, knowing the list of words and nothing else would be rather pointless now wouldn’t it (c.f. my dinosaur example), but come on, really what teacher is going to teach the list of meaningless words and offer no explanation as to what each one means!  The mnemonic song helps to provide cues to recall at least the names of the prepositions.  From there, hopefully the meaning will become associated with the names through another process, and their application should be repeatedly practised.  The mnemonic is one tool, and one used to provide internal cues so as to aid recall – it is not the sum total of all that is to be learnt about prepositions.

* Mnemonics do not preclude meaningful understanding.

Ball described them as ‘disposable crutches’ – we use them until we don’t need them anymore.  This is also echoed in what Hattie and Yates say about mnemonics.  To remove them would be like removing the lower rungs on a ladder, arguing that because our goal is to reach the top, we could save time by missing out those earlier steps.


Is this really what we think education should look like?

  • Memorisation is learning.
  • Inflexible knowledge is a necessary first step on the path to flexible knowledge.
  • Distributed practice builds storage strength.
  • Mnemonics do not preclude meaningful understanding.

By now it may become apparent that when and where we decide to use different techniques and strategies to promote long-term recall is dependent upon the very nature of what we want to teach, and what we want pupils to take away from our teaching.  Orlin is worried that mnemonics replace ‘meaningful’ learning.  In some of the examples above there really isn’t any more meaning to take, beyond just knowing, say, which line on the stave refers to which musical note.  Are there alternative things we may wish pupils to learn, then, where mnemonics and ‘raw rehearsal’ really could promote the accumulation of meaningless, isolated facts?

We need to build taxonomies of our subject content to help answer this question.  These taxonomies can then help us in deciding how to communicate and assess the learning of different kinds of knowledge.  In the final post in this series we’ll look at an example Orlin uses where the knowledge to be learnt is arguably quite different from some of my examples above.  Orlin’s recommended alternatives to raw rehearsal and mnemonics are ‘repeated use’ and ‘building on already known facts.’  I will highlight a misconception in the notion of ‘repeated use,’ as well as one of the risks associated with ‘building on known facts’ as a strategy on its own, while adding a new layer of nuance to the use of mnemonics to promote ‘meaningful’ learning.

About Kris Boulton

Teach First 2011 maths teacher, focussed on curriculum design.
This entry was posted in Uncategorized. Bookmark the permalink.

6 Responses to Does memorisation get in the way of learning? – Part 3

  1. Pingback: Does memorisation get in the way of learning? – Part 4 | …to the real.

  2. mrdardy says:

    Almost done with these posts – been taking them in order and had read (and enjoyed) Ben Orlin’s original post. I want to comment on the mnemonic conversation. Your mention of SohCahToa is important. There is nothing conceptual in nature about how these ratios are defined. They just are what they are. Another math mnemonic that is used is FOIL. Here, I’d argue that the mnemonic gets int he way of learning because there is a procedural logic behind distributing and it’s incorrect to train kids to think that there is only one order to do this.

    • Kris Boulton says:

      Thanks for the comment.

      There’s plenty that’s good in the points Ben Orlin is trying to make, and it’s fair to say that there is very often no (maybe never any…?) conceptual underpinning in the mnemonics we use. Mnemonics essentially help us to ascribe some kind of connected meaning to things that can otherwise be meaningless. For example, you could recall the order of the planets if you knew their average surface temperatures. While Venus breaks the rule, knowing why only imbues a person with further understanding. This way of remembering the order of the planets has two advantages over an acrostic mnemonic (like ‘My Very Easy Method Just Shows Us Nine Planets); it provides a much richer understanding of the properties of the planets, and how and why they vary.

      There are two problems with this though. The first is that there’s still a bunch of stuff to remember, and the second is that working out the order from recall of the temperatures would take longer than the nice, easy mnemonic that most of us still probably use even in adulthood.

      The same is true I think of Orlin’s example about Maryland in the civil war. Yes, his version lends itself better to deeper understanding – of course it does! Who could deny that. But to say therefore that it is simply *better* per se than using a mnemonic, or that mnemonics shouldn’t be used at all, or that they’re damaging, is incorrect. In fact, in the deepest of ironies, I just typed out ‘Massachusetts’ and went back to change it to ‘Maryland’ because I remembered that ‘Marriage is a union!’ As I wrote this paragraph, I couldn’t actually remember what the facts were that Ben wanted us to recall… now I remember that ‘Marriage is a union,’ I recall that it was about Maryland being a part of the union during the civil war, and that has served as the cue for me to recall that Lincoln worked hard to keep it as part of the union because, had it ceded, Washington would have been surrounded. This was one of the points that I tried to make in the series, that sometimes mnemonics serve as cues, or an ‘in’ to recall the more related things we want people to know – they don’t have to replace it.

      My issue with the suggestion in Orlin’s article was that, due to the advantages of conceptual understanding, mnemonics should therefore never be used. This view seems to arise for a few reasons. One seems to be the misconception that ‘meaningful’ or conceptual understanding automatically leads to strong memory recall, as if there is nothing at all to work to recall if we’ve learnt it in that way! Another is the idea that learning a mnemonic necessarily means a person then cannot go on to understand something at a deeper level.

      Rather than the ‘always or never’ mnemonic dichotomy, I think it should be judged on a case by case basis, and the litmus test maybe should be ‘will a mnemonic aid recall in this instance?’ I’ve never been a fan of FOIL for example – I don’t think it’s a great mnemonic, and a simple visual aid plus practice is usually more than enough. James Tanton is another teacher who seems to display a dislike for things like mnemonics and ‘memorising’ things (which I sometimes find ironic, given the vast range of things he shows us he’s clearly memorised!) Like Ben though, he does also have some excellent explanations for mathematical ideas. His explanation for expanding brackets looks great. I’ve only skimmed the PDF, but it looks like it’s sequenced and structured in a way that would lead to a great round of explanation and practice, with no need for any mnemonics later:

      Maybe you’re even write to criticise it for unnecessarily and deeply embedding misconceptions. This is an area where we need to be careful – sometimes I think it might be okay to learn to ‘do something,’ even though there might be more exotic cases where it doesn’t apply, then, learn about those cases later. This might be necessary as it could be the only way of bridging towards a very complex concept. FOIL, probably isn’t one of those times though, and it probably embeds that misconception unnecessarily. That said, there *is* a better order to run the process, as it gives you the expression in the convention format at the end – these are things that should all be discussed during the teaching process though.

      SohCahToa on the other hand… heck, even I still use it, all the time! Knowing that certainly hasn’t impeded my ability to delve into a deeper conceptual understanding on trigonometry though; not being taught those conceptual ideas is what got in the way!

      I guess in summary, I don’t think that it’s always one or the other – a tack that too often is put forward in education debate. I think mnemonics, can, do, and should serve a powerful purpose in our teaching, but that doesn’t mean I think they should be used badly, or when they’re not effective.

      What do you think?

  3. mrdardy says:

    Wow – thanks for the thoughtful reply. You are certainly correct that there is a false dichotomy here. Almost all of them ultimately are, I suspect. Looking back at your post, the elephant picture related to the word BECAUSE reminded me of a silly story. One day at school I was walking to lunch and a student stopped me. She said, I just came up with a mnemonic to remember how to spell my name. This was a high-school student named Jeanine. She told me her mnemonic was ‘Just Eating Aardvark Noses In New England’ I think in some ways this incident summarize how I feel about mnemonics. That being said, as I mentioned in my last post, there are certain ones that are certainly helpful and don’t get in the way of any conceptual understanding. I think that we both would agree that we don’t want to interfere with understanding the underlying ideas of any of our studies. I would also argue – probably too optimistically – that concentrating on understanding helps with quicker recall of simple facts. I’d lean in that direction with the Maryland conversation. Certainly no harm in having a quick recall tool, but recognizing the idea here also helps with quick recall. I’d say that the litmus test you propose is a reasonable one, but I’d want my litmus test to be ‘Will this mnemonic obscure an important underlying principle?’ In he case of SohCahToa the answer is no. In the case of the mnemonic about the order of planets, the answer is no. When the answer is no, I say go ahead with the shortcut. If the answer is yes, work hard to go around it.

  4. Pingback: I ♥ rote learning | David Didau: The Learning Spy

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s