Words before you listen: why preparing helps with podcasts

What pre-teaching is

In school, this is called pre-teaching. Before the class listens to a text, the teacher prepares them on the topic and clears up the vocabulary. The Lehrerfortbildung Baden-Württemberg (teacher training) names among the aims of this pre-listening phase to tune the students in to the text and to “relieve the vocabulary beforehand”, so that they can then listen with focus.

If you listen to English podcasts on your own, nobody does this for you. You hear the episode cold, stumble over a word and look it up afterwards, if at all. Pre-teaching for self-learners turns the order around: first the few words you would get stuck on, then the episode.

Why an unknown word disturbs you more when listening than when reading

When you read, you stop, read the sentence again and check a dictionary. When you listen, the speaker simply carries on. While you are still wondering what garrison might mean, the next sentence is already playing, and you miss that one too.

In 2000, Goh evaluated what learners fail at when listening. Among the five most common difficulties is that they miss the next part while they think about the meaning. Another is that they do not recognize words they actually know, because spoken they sound different. A 2023 study that summarizes Goh’s list observed such lapses mostly with longer listening texts, and a podcast episode often lasts an hour.

With film, you at least have the picture: research on vocabulary in films and series expressly names facial expressions, gestures and lip movements as support. With a podcast, you only have the voice.

Why native speakers also sound so fast is in the guide Understanding fast English.

How many words you need to know for a podcast

Research measures this by what percentage of the words of a text you know, and calls it coverage. Words are counted in word families: a base word with its forms and derivations, so decide, decided and decision together.

• 98%: A Danish study with 115 advanced learners estimated that for a listening test of this kind you need to know about 98% of the words (Stæhr 2009). For spoken English, Nation (2006) arrives at 6,000 to 7,000 word families.
• 90 to 95%: In 2013, van Zeeland and Schmitt tested listening specifically. For most participants 90% was already enough, but among the non-native speakers, understanding then varied a lot. From 95% they understood relatively well and more evenly. For that, 2,000 to 3,000 word families are enough. So the estimates differ; which value is right depends on the listening text and on the study.
• Films and series: According to Ha (2022), you need 3,000 word families for 95% of the words in film and series dialogue, and 5,000 for 98%.

This is what 95% looks like in a sentence. It has 20 words, and you do not know one of them:

The garrison held out for three weeks, but by then the city was starving and the king had already fled.

word you knownew word you clear up beforehand

At 95%, you meet a word like that in every twentieth word, at 98% in every fiftieth. Ha calls that a huge difference for three percentage points.

The example sentence shows two more things. First, the missing word is the most important one: if you do not know garrison, you do not know who held out for three weeks. What you lack in words is mostly the rare ones that carry the topic of an episode, and in every episode you lack different ones. Second, these studies count single word families. Held out falls through the grid: you know hold and out, but the count does not tell you that both together mean “to keep going”.

What research shows, and what it does not

People who know more words tend to understand more

In 2022, Zhang and Zhang evaluated more than 100 studies with almost 21,000 learners in total (a meta-analysis). Between English vocabulary and listening comprehension they found a moderate correlation of about r = .56. The Danish study by Stæhr (2009) points in the same direction: vocabulary was strongly linked to listening comprehension.

That is an association, not an effect. The studies do not show that learning words improves listening comprehension, and they say nothing about a step directly before listening. Of both, I only checked the abstract, not the full text.

Learning vocabulary beforehand: the evidence is thin

Aldukhayel (2023) randomly assigned 93 Arabic-speaking English learners to four groups. One learned only vocabulary before a listening text of just under five minutes, one got a preview of the content, one got both, one got nothing. Only the vocabulary group improved. The groups were small, the test was short, and the advantage showed mostly among the advanced learners. The study itself calls the samples small.

Chang and Read (2006) found the opposite with 160 students in Taiwan: vocabulary instruction helped least, background knowledge on the topic helped most. Beginners said they had not yet known the words securely. Advanced learners found that the list had distracted them while listening.

What follows from this

The studies do not give you a “vocabulary beforehand helps”, and Wordspoiler does not promise results. Both experiments took place in the classroom, with short listening texts instead of podcasts. For your everyday life, though, quite a lot speaks for a short list with the really important passages: it distracts less while listening than a long one. Knowledge of the topic helps at least as much as the words.

Which words you should clear up beforehand

Not every unknown word is worth it. Choose like this:

• Rare, but carrying. Words without which you do not know what it is about, like garrison above. A word that falls once in a subordinate clause, you can skip.
• Phrases before single words. A fixed phrase like hold out (“to keep going”) is harder to recognize when listening than a single word, because you know every part and still not the whole. That is exactly why Aldukhayel included idioms and fixed combinations in the list.
• Names and technical terms of the topic. A name you have never seen written, you hardly recognize spoken.
• Nothing you can guess. If the sentence gives away the meaning, you do not need the word beforehand.

Look at each word in its sentence from the episode, so that you recognize the passage when you hear it. The translation alone is not enough for that. Note roughly where in the episode it occurs, say after ten minutes. And listen to the pronunciation: Aldukhayel’s vocabulary group also got the list as an audio recording, and Goh’s learners often did not recognize words they actually knew when spoken.

A routine for 10 minutes before the episode

This is how it works without an app, with a transcript or the episode notes:

1. Choose an episode whose topic you know or that interests you. Knowledge of the topic helped most in Chang and Read, but that too is only one study. Which podcasts fit which level is in our overview English podcasts by level.
2. Get the text. Many learner podcasts have a transcript on the episode page. For 6 Minute English from BBC Learning English, it is under every episode, plus a box with six vocabulary items. Since 2024, Apple Podcasts on iPhone and iPad automatically generates transcripts for English episodes. Without a transcript, you have the episode description, which often names topics and names.
3. Skim the text and mark 10 to 20 passages. Phrases first, then rare words, then names. Stop when you reach 20.
4. Look them up, with pronunciation. An online dictionary with an audio example is enough. Write down the meaning in your own language and the sentence from the episode.
5. Note the passage roughly. “At the start” or “shortly before the end” is enough.
6. Read the list aloud once, then listen to the episode. Do not stop to look things up. When a word from your list comes up, you usually notice it by yourself.
7. Two minutes afterwards. Which words did you recognize? The others you keep for practice.

The real work is in steps 2 to 4. The first time, it takes more like 15 minutes, and for some podcasts you find no text at all.

Where Wordspoiler saves you work

Wordspoiler takes over steps 2 to 5: the app picks out the passages and shows each with its sentence from the episode and the approximate time, for example “approx. 11:30”. It shows 20 to 30 passages instead of 10 to 20, because you swipe away the ones you know in seconds: you only learn what is left. What you swipe away does not come up in later episodes either. Then you listen to the episode as usual in your podcast app.