Creating audio files for frenchteacher.net

I've recently had a bit of fun creating text-to-speech audio files for some of the listening tasks on frenchteacher. Up to now, the listening materials on the site have come in three forms:

Read aloud tasks from Y9 up to A-level. The Y9 tasks are in the form of narrow listening exercises, each with four short paragraphs featuring repeated language patterns 'input flood' style). Each task consists of a teacher sheet for reading aloud, and pupils worksheet. The Y10-11 tasks are passages to be read aloud by the teacher with accompanying worksheet exercises to print off. Some are pitched at Higher GCSE, some at Foundation. The A-level tasks have a passage to read aloud, with an accompanying worksheet. Topics are in line with exam board specifications.
Audio listening. These are mainly exercises linked to Audio-Lingua listening clips are and free to use on the Samples page.
Video listening. These are found largely from Y10 up to A-level and are worksheets linked to online video clips, for example from YouTube and French news channels. Most listening resources on the site are in this form, since authentic video is rare in text book courses and expensive when accessed by, for example, This is Language.

So what I have done is create audio files for the read aloud tasks in Y10-11 and A-level. I have done this using the text-to-speech facility within Microsoft Windows, helped along with a Microsoft app called Text to Speech and MP3.

You can find the app here:

Get text to speech & mp3 - Microsoft Store en-GB

It is partly free, because if you want to save MP3 files (which I did) you pay around £9 for an add-on. As long as your target language is enabled on your PC (French and Spanish are built in) you can just import your chosen text, or type it in, then choose a voice to read it aloud. There are three voices for French, two female, one male (Microsoft Julie, Microsoft Hortense and Microsoft Paul). Press play in the app and you hear the text read aloud. There is a choice of speeds, but I found normal (100%) is fine.

Below is what a screen looks like (from the app site):

I have no idea how you would do this on a Mac, I'm afraid.

As is usually the case with text-to-speech, the pronunciation is very good, but the intonation sounds artificial. To render it less artificial, you need to add quite a lot of extra commas to the written text so that the text is read in meaningful phrases. The result is not entirely natural, but it's pretty good and I would be happy to use it occasionally.

Once you have tested the file, you then save it as an MP3. In my case, I was then able to upload it my site, each audio file indicated by a little headphone icon I found on Pixabay. the screen shot below is from the Y10-11 page.

So I now have around 25 audio files which can be played by the teacher as an alternative to reading aloud. Note that the idea is not that students use these files themselves, since students don't have their own log-ons. (Teachers are asked not to share these with pupils.)

My hope is that these will get some use, as an occasional alternative to the teacher doing the reading.

If you want to hear text-to-speech used effectively in an online resource, try TeachVid or the new SentenceBuilders site, both from Martin Lapworth.

What is skill acquisition theory?

For this post, I am drawing on a section from the excellent book by Rod Ellis and Natsuko Shintani called Exploring Language Pedagogy through Second Language Acquisition Research (Routledge, 2014). Skill acquisition is one of several competing theories of how we learn new languages. It’s a theory based on the idea that skilled behaviour in any area can become routinised and even automatic under certain conditions through repeated pairing of stimuli and responses. When put like that, it looks a bit like the behaviourist view of stimulus-response learning which went out of fashion from the late 1950s. Skill acquisition draws on John Anderson’s ACT theory, which he called a cognitivist stimulus-response theory. ACT stands for Adaptive Control of Thought. ACT theory distinguishes declarative knowledge (knowledge of facts and concepts, such as the fact that adjectives agree) from procedural knowledge (knowing how to do things in certain situations, such as understand and speak a language).

What is the natural order hypothesis?

The natural order hypothesis states that all learners acquire the grammatical structures of a language in roughly the same order. This applies to both first and second language acquisition. This order is not dependent on the ease with which a particular language feature can be taught; in English, some features, such as third-person "-s" ("he runs") are easy to teach in a classroom setting, but are not typically fully acquired until the later stages of language acquisition. The hypothesis was based on morpheme studies by Heidi Dulay and Marina Burt, which found that certain morphemes were predictably learned before others during the course of second language acquisition. The hypothesis was picked up by Stephen Krashen who incorporated it in his very well known input model of second language learning. Furthermore, according to the natural order hypothesis, the order of acquisition remains the same regardless of the teacher's explicit instruction; in other words,

The 2026 GCSE subject content is published!

Two DfE documents were published today. The first was the response to the consultation about the proposed new GCSE (originally due in October 2021) and the second is the subject content document which, ultimately, is of most interest to MFL teachers in England. Here is the link to the document. We are talking about an exam to be done from 2026 (current Y7s). There is always a tendency for sceptical teachers to think that consultations are a bit of a sham and that the DfE will just go ahead and do what they want when it comes to exam reform. In this case, the responses to the original proposals were mixed, and most certainly hostile as far as exam boards and professional associations representing the MFL community, universities, head teachers and awarding bodies are concerned. What has emerged does reveal some significant changes which take account of a number of criticisms levelled at the proposals. As I read it, the most important changes relate to vocabulary and the issue of topics

Paul30 April 2021 at 08:49
Thanks for the post. I've also been experimenting with TTS and found that the quality of the modern voices is incredible. I've tried sites like freetts.com and www.texttomp3.online which create mp3 files from text using TTS but the voices are average. The best is to use the "Natural" voices in Microsoft's Immersive Reader. The easiest way to use this is to put your text into Microsoft Word Online (Microsoft's freemium answer to Google Docs) at office.com . This includes the Immersive Reader with the "Natural" AKA "Neural" voices. By setting the correct "Proofing language" you can even mix languages (but not accents) in the same document. The documents can be shared with students and render well on all the devices that I've tested. To find "Immersive Reader" in Word Online click "View" > "Immersive Reader".
Take things to the next level with SSML and craft conversations using this tool speech.microsoft.com/audiocontentcreation - it lets you create conversations with different voices. You can record your computer audio using the free Audacity program.

Language Teacher Toolkit: Steve Smith's blog

Search This Blog