Photo by Melanie van Leeuwen on Unsplash

There’s an argument that comes up occasionally that, in addition to being worth studying for the sheer education of it, Latin and Greek have utilitarian side benefits — it gives pupils skills and understanding that can be used in a variety of other subjects.

This argument is a frequent refrain of Boris Johnson:

Hard as it may be to believe, one of the things that gives privately-educated children the edge is their knowledge of Latin. I don’t just mean in the obvious senses — their grasp of basic grammar and syntax, their understanding of the ways in which our world is underpinned by the classical world, their ability to read Latin inscriptions. I mean there is actually a substantial body of evidence that children who study Latin outperform their peers when it comes to reading, reading comprehension and vocabulary, as well as higher order thinking such as computation, concepts and problem solving.

The first thing to understand is the role of confirmation bias. Johnson is an excellent example of a man who became successful, rather enjoyed Latin and is keen to draw a connection between the two. There’s nothing really wrong with that — but we need to be wary when people start talking about ‘evidence’ that the thing they like also happens to be empirically good for you.

So what is the actual evidence on Latin?

Very roughly, there are mixed results in a few areas and some definite negatives that undermine the claims that Latin is education pixie dust.

The first thing to investigate is if Latin can really do work on skills that seem outside its domain. Haag and Stern (2003) have published two studies on the subject. In the first they found no evidence that it helps out with maths or verbal and non-verbal reasoning skills — there was no difference between German students who were taught English or Latin as their foreign language. In 2003 they published another study showing “no significance differences (all ps > .20) in deductive and inductive reasoning or text comprehension” when comparing students with no Latin, two years of Latin, and four years of Latin.

So un-encouraging there, but what about language skills? Johnson talks about Latin being a “giant universal spanner for other languages”, and this kind of argument is common, Llewelyn Morgan argues that the language skills built learning Latin have wide applications:

A child who has been introduced to Latin has the tools and (just as importantly) the confidence and interest to take on other languages. And we are not just talking about the Romance languages that derive from Latin. German is not a Romance language, but it is, like Latin, an ‘inflected’ language, and the difference between der, den and das is self-explanatory when a child has already played ‘declension cricket’ with bellum or mensa.

This turns out to have more basis in fact. Indeed in Haag and Stern’s 2003 study the group of German-speakers with four years Latin exposure were better able to “detect grammar mistakes in a German text and construction of complex sentences by combining shorter ones” than those who studied it for 2 years or not at all.

Here is the obvious question that raises: If A makes B easier but A also takes time and effort, can that time be spent directly on B and get better results? How much Latin do we really need? Can we dump all that history and culture stuff and get the same benefit?

The answer seems to be yes — Holmes and Keffer (1995) found that a simple six week period using a computer program to teach Latin and Greek words and roots could boost the mean score of students sitting the American SAT exam by 40 points. This kind of result is why classicists should be a bit careful promoting the practical benefits, what the study is actually saying is you can get substantial benefits by carving the Classics curriculum up for parts and incorporating a shallow version of it into the English curriculum.

In fact a 1970s study of a Latin program in US elementary schools found that both Latin and modern languages courses were increasing English vocabulary and comprehension compared to students doing neither. Interestingly the Latin students had only been studying for one year while the foreign language students had four (Masciantonio, 1977, 378). This either implies that language instruction quickly has diminishing returns or that Latin was just that much better — either way showing a shallow take on Classics would probably give most bang for buck.

Reinforcing that the Latin education with the most spillover benefit would probably look quite different to how it’s usually conceived, Masciantonio argued from a review of a studies on the impact of Latin instruction that “it is not the traditional grammar-translation approach that has yielded the positive results indicated in this article; rather it is programs that involve radical reform of both curriculum content and instructional strategies” often making “extension of the English verbal functioning a specific goal” (Masciantonio, 1977, 382).

If you’re looking at it from a utilitarian point of view it’s not enough to show that learning Latin is a good use of time to accomplish other tasks, you have to ask if it’s the best possible use of time. Hagg and Stern found that students studying Latin rather than French were disadvantaged in learning Spanish — with the French group being notably superior in vocabulary and use of grammar rules. Deploying the “universal spanner” of Latin was in fact causing students to make mistakes due to superficial similarities between Spanish and Latin words. For a companion to learning a modern romance language the best answer seems to be to learn another one at the same time. For all the talk of Latin and Greek as ‘root’ languages there are significance differences between their structure and modern languages. After all, the mantra that Latin has a lot to tell us about our own language led to generations of pedants who would rather we torture words than split infinitives.

So if you want to learn the most effective way to incorporate Latin and Greek into English education, there’s a lot of stuff out there — it’s less clear this works as an argument for Classics education as it stands.

Latin Supremacy in Latin Advocacy

The idea that Latin gives unique benefits is also used as part of an argument that we live in a fair society — so the arguments being deployed need to be looked at more closely.

Latin education in the UK is tied up with historical elitism in the education system — Morgan argues that modern Classics education has moved past this:

Note here that no one, at least no one who knows what they’re talking about, is calling Latin ‘exclusive’ or ‘elitist’ anymore. Latin is classless. Our local experience, in a Latin Teaching Scheme run out of the Oxford Classics Faculty in collaboration with local state schools, is of a subject school children find entirely accessible, and enjoyable: the drop-out rate for sessions that take place on a Saturday is surprisingly low, and the demand growing.

This seems fair enough, it’s hard to argue that any particular subject is inherently elitist— and making that learning available to more certainly can’t be. However if you read between the lines in Latin advocacy you can see an old world trying to justify its existence. Dr Peter Jones says in support of Classics For All:

“[W]e know that those who have studied the ancient languages are never, in fact, short of job-offers. A top asset-manager recently told me that his firm always employed classicists: they sold more. If Richard Dawkins is right, that is because ‘what Classics has always done is just teach people how to think.”

To be fair to Jones that’s an aside from his argument, but if you join that small point up with Johnson’s comment about Latin being the real cause of the success of the privately educated you can see the same argument at work. “Latin leads to success” is a reflexive self-justification that the extreme successes of some classicists is more substantial than just passing a cultural shibboleth. If Latin really provides skills then a society where Latin-speakers prosper is meritocratic, not elitist. On the other hand, if there’s no evidence that Latin is giving benefits that can’t be picked up in other ways we might have to ask awkward questions about where the elite come from.

Behind the idea that Latin improves proficiency in other areas lurks the older idea that Classics is true education, of which all other areas are inferior imitators. It is bedrock, fundamental, “the Maths of the Humanities”. Some defences treat classicists almost as the guardian of the Old Ways who will save us from the nasty, modern, English-hating English teachers. The think tank Parliament Street states this more plainly than most:

In the past at traditional schools it was not uncommon for children to have the same teacher for English and Latin in their early years, and there is much to be said for this, given the apparent reluctance of so many English teachers to teach their own language.

So mixed in with Latin advocacy is a effort to defend the educational structures and outcomes of the past — but alongside that you’ll find enthusiastic former students and current teachers. These people have a good-hearted enthusiasm of something they love and a desire to pass that on to the next generation. What could possibly be wrong with that?

The thing to remember everyone thinks that the thing that formed them is special and that other people would benefit if they had more of it in their lives - be it Latin, Dance or Ice Hockey. If the utilitarian argument falls apart Classics has to compete with the multitude of other life changing opportunities available.

So yes, Latin should be compulsory. But so should programming. And drama. And Mandarin. They should learn special relativity in the morning and engineering in the afternoon so they can invent more hours in the day to learn all the things that are absolutely required to understanding the world.

In the event they can’t pull this off, we might have to accept that people leave school hopelessly ignorant of many wonderful things.


Haag, Ludwig, and Elsbeth Stern (2003). “In Search of the Benefits of Learning Latin.” Journal of Educational Psychology 95, no. 1 : 174–78. doi:10.1037/0022–0663.95.1.174.

Holmes, C Thomas, and Ronald L Keffer (1995). “Computerized and on Greek Verbal Method Words : Scores to Teach Latin Effect.” English Today 89, no. 1 : 47–50.

Masciantonio, R. (1977), Tangible Benefits of the Study of Latin: A Review of Research. Foreign Language Annals, 10: 375–382. doi: 10.1111/j.1944–9720.1977.tb02999.x