Oxford University Press's
Academic Insights for the Thinking World

How do people read mathematics?

At first glance this might seem like a non-question. How do people read anything? All suitably educated people read at least somewhat fluently in their first language – why would reading mathematics be different? Indeed, why would mathematics involve reading at all? For many people, mathematics is something you do, not something you read.

But it turns out that there are interesting questions here. There are, for instance, thousands of mathematics textbooks–many students own one and use it regularly. They might not use it in the way intended by the author; research indicates that some students–perhaps most–typically use their textbooks only as a source of problems, and essentially ignore the expository sections. That is a shame for textbook authors, whose months spent crafting those sections do not influence learning in the ways they intend. It is also a shame for students, especially for those who go on to more advanced, demanding study of upper-level university mathematics. In proof-based courses it is difficult to avoid learning by reading. Even successful students are unlikely to understand everything in lectures – the material is too challenging and the pace is too fast – and reading to learn is expected.

Because students are not typically experienced or trained in mathematical reading, this returns us to the opening questions. Does this lack of training matter? Undergraduate students can read, so can they not simply apply this skill to mathematical material? But it turns out that this is not as simple as it sounds, because mathematical reading is not like ordinary reading. Mathematicians have long known this (“you should read with a pencil in hand”), but the skills needed have recently been empirically documented in research studies conducted in the Mathematics Education Centre at Loughborough University. Matthew Inglis and I began with an expert/novice study contrasting the reading behaviours of professional mathematicians with those of undergraduate students. By using eye-movement analyses we found that, when reading purported mathematical proofs, undergraduates’ attention is drawn to the mathematical symbols. To the uninitiated that might sound fine, but it is not consistent with expert behaviour; the professional mathematicians attended proportionately more to the words, reflecting their knowledge that these capture much of the logical reasoning in any written mathematical argument.

Another difference appeared in patterns of behaviour, which can best be seen by watching the behaviour of one mathematician when reading a purported proof to decide upon its validity (see below). Ordinary reading, as you might expect, is fairly linear. But mathematical reading is not. When studying the purported proof, the mathematician makes a great many back-and-forth eye movements, and this is characteristic of professional reading: the mathematicians in our study did this significantly more than the undergraduate students, particularly when justifications for deductions were left implicit.

This work is captured in detail in our article “Expert and Novice Approaches to Reading Mathematical Proofs”. Since completing it, Matthew and I have worked with PhD and project students Mark Hodds, Somali Roy and Tom Kilbey to further investigate undergraduate mathematical reading. We have discovered that research-based Self-Explanation Training can render students’ reading more like that of mathematicians and can consequently improve their proof comprehension (see our paper Self-Explanation Training Improves Proof Comprehension); that multimedia resources designed to support effective reading can help too much, leading to poorer retention of the resulting knowledge; and that there is minimal link between reading time and consequent learning. Readers interested in this work might like to begin by reading our AMS Notices article, which summarises much of this work.

In the meantime, my own teaching has changed – I am now much more aware of the need to help students learn to read mathematics and to provide them with time to practice. And this research has influenced my own writing for students: there is no option to skip the expository text, because expository text is all there is. But this text is as much about the thinking as it is about the mathematics. It is necessary for mathematics textbooks to contain accessible text, explicit guidance on engaging with abstract mathematical information, and encouragement to recognise that mathematical reading is challenging but eminently possible for those who are willing to learn.

Featured image credit: Open book by Image Catalog. CC0 1.0 via Flickr.

Recent Comments

There are currently no comments.