Why you shouldn’t mark spelling on a dictée

It’s common practice to give a French dictation, or dictée, and mark it for accuracy. Did the students write down what you said? Did they reproduce on their pages what was written on yours?

One of the look-fors in terms of accuracy is the student’s spelling. But it shouldn’t be.

It doesn’t assess the listening skill that dictation actually elicits from a student.

Dictation in a French class primarily assesses auditory parsing. This might be defined as the ability to know what you’ve heard, to break down the stream of sounds into words. For example, if you turn on the radio in French and listen to the news or a new song, your comprehension depends in large part on your ability to parse what you hear. Can you tell which words they’re using? (Even if you don’t know what all of them mean?)

This is a different skill from spelling, which consists of knowing certain patterns of language, memorizing exceptions, and understanding the grammatical concepts of gender, pluralization, and edge cases of liaison.

My own experience tells me that these are two very different skills. My French spelling is excellent, probably because I majored in English and French linguistics and have worked as a proofreader in both languages. I can distinguish the passé simple from the subjonctif de l’imparfait, and if you show me an incorrect spelling, I can guess reasonably well whether it’s a student’s mistake or the word as Molière wrote it.

Auditory parsing, on the other hand, is probably my weakest skill in French. When I visited Paris, I often had trouble making out what someone said. Once, I embarrassingly failed to understand a waiter’s question « Dîner ? » I know that basic word and can spell it. But my brain didn’t break down the sounds I heard into the word that I knew.

But there are more striking examples than my own. An illiterate person can have no spelling expertise but fantastic auditory parsing, and a hearing-impaired person can have no auditory parsing expertise but fantastic spelling. They simply aren’t the same operation.

The factors that influence auditory parsing include words per minute spoken, the ability to see the speaker’s lips, background noise, and vocabulary size. All but the last have no connection whatever to spelling, and even vocabulary size isn’t the best predictor of spelling.

The factors that influence spelling, meanwhile, are equally unrelated to auditory parsing. In fact, in French, you can’t hear most of the letters in a word, and — although it isn’t as irregular as English — the relationship between spelling and pronunciation is not always predictable. You don’t hear letters; you hear words. And for all the silent letters in French, that’s not even a perspective shift, just an observation.

It’s not your ears that help you choose a spelling; it’s your background knowledge of grammar and vocabulary.

“But it makes so much sense to bundle spelling with auditory parsing by marking both in a dictation. That’s why it’s an effective assessment tool.”

As teachers, we certainly do like the idea of assessing more than one thing at once. But these things have to be loosely related; the best is if they work in tandem, if they influence each other in some way. When that’s not possible, as with spelling and listening, you can at least give two separate marks. That’s the only way to make it useful to a student. Otherwise, if a student’s listening improves and they write down the correct word, but they still spell it incorrectly, their mark will fail to reflect their gains.

Even better, though, is to use an assessment tool that doesn’t confound the two skills at all. You might wonder how you can test spelling without using dictation — you can’t exactly show them the written word, and how else do they know which words to spell? But that’s exactly my point. You can’t assume that “they know which words to spell” when giving a dictation task. You can’t presuppose it; you’re trying to assess that very fact at the same time.

Some ideas you probably already use to assess spelling in a more independent way:

• Mark spelling when they write and edit, particularly when they do so by hand. In that situation they know which words they are trying to spell.
• Do spelling fix-up activities using examples of incorrect spelling.
• Have students predict spelling when given the context and rules that actually determine it (e.g. “A girl texts her new friend to say she’s athletic. Does she say « Je suis sportif » or « Je suis sportive » ? Does she type an f or a v?”).
• If you do dictation to assess spelling, go very slowly and make sure the speaker’s lip movements are clearly visible, even exaggerated. If possible, let them listen as many times as they like. You can (and should) take out those aids when, on the other hand, you do dictation to assess auditory parsing.
• If you do dictation to assess both at once, at least mark it in two rounds. First check that they got the right words, spelled correctly or no. Then have them do a second pass where they fix their own spelling.

Some ideas to assess auditory parsing on its own:

• Have students listen to a dictation and orally repeat it, rather than writing it down.
• Have students listen to a dictation at home and record themselves repeating it back.
• Have students listen to a dictation and answer very simple comprehension questions that require gaps to be filled in (e.g. « Le canard s’est envolé. — Qui s’est envolé ? »).
• Have students create a conversation where they hear a very simple prompt and have to give an appropriate reply, perhaps selecting from a word bank (e.g. they hear « Bonjour ! Ça va bien ? » and choose between « Oui merci », « Bonjour », and « Comment ça va ? »).
• Have students listen to a dictation and write it down, but don’t mark the spelling.

Using these tools you’ll soon be assessing what you meant to assess. The students will feel it’s fairer, and you’ll have a better understanding of where they are.

Appendix: It might be that the use of dictation for spelling tests is carried over from English education, or even French education in France. I remember many spelling tests in my native English that were given by dictation.

But that’s a different case. Children parse their native language nearly perfectly already (all things being equal), so you know they’ll catch every word just fine and all you’re assessing is spelling.

Given that, a teacher might reply: “My students’ listening and speaking skills in French are great. They always understand what I say orally, so I don’t need to worry about that factor in dictation.”

Maybe, but maybe not. I’ve had many an immersion student who can speak quickly on their feet, and who can respond appropriately to what they hear. But appearances can be deceiving. More often than not, I find they can make up for weak auditory parsing using strong conversational skills. That is, they don’t always know what they heard and didn’t hear, but they can do a pretty good job using their comprehension of the sentence as a whole to avoid confronting that.

So in reality, many of these students still need to be assessed and taught in auditory parsing — at least until they’re at a very high level. When I do dictation, they don’t get 100% of the words, spelled right or no. So I don’t use my dictation exercises with them to test spelling — just parsing. Then I assess spelling in other ways.

I wonder what you think?