How does speed listening work?

Question

I've recently gotten into listening to podcasts. Over time, as I get accustomed to the speaker's voice, I'm able to increase the speed of the podcast to as high as 3x speed. It still feels "normal" to me, as if the speaker is having a normal-speed conversation with me.

When I turn down the speed to 1x, it sounds unbearably slow. For non-speed-listeners, listen to something at 0.5x or 0.25x, and you'll know what 1x sounds like to speed listeners. It's as if speed listening shifts up a listener's range of acceptable speed, so that lower speeds are no longer in that range.

I know there are plenty of other speed listeners who do the same thing (just Google "speed listening" to see for yourself). Yet, others have told me that at this speed, podcasts are basically unintelligible.

Thus, I have a handful of related questions:

1. What occurs in the brain when a person speed listens?

   a. Why does the podcast feel "normal" at such fast speeds?

   b. Is there a fundamental/biological difference between those who can speed listen and those who can't, or is it a learned ability?

2. Is speed listening detrimental to listening comprehension?

Answer 1

Your question made me think of JAWS, a screen reader for the blind. I have worked with visually impaired people for a while and I have always wondered how on earth they can understand the speech produced by JAWS given the sheer high speech rates they apply on their gadgets.

Indeed, people with peripheral vision loss may learn to understand spoken language at rates of up to about 22 syllables per seconds, while the rate in normal listeners is about 8 syllables per sec (Dietrich et al., 2013). These numbers are based on speech understanding, as opposed to simply hearing it.

Hence in answer to your question

Is speed listening detrimental to listening comprehension?

I would say no, at least not in above study, as speech understanding there was tested in a speech comprehension test, which is different than just hearing the speech.

Then your questions

What occurs in the brain when a person speed listens?
a. Why does the podcast feel "normal" at such fast speeds?
b. Is there a fundamental/biological difference between those who can speed listen and those who can't, or is it a learned ability?

In case of blind people it has been shown that they activate their visual cortex when speed reading. This, because the deafferented area is recruited for other modalities, such as touch and hearing. This is probably not what you are after in your question though. In effect, I doubt gross effects on brain activity occurs in trained, normally sighted subjects and indeed a study where both sighted and blind individuals were trained on fast-speech understanding does not report on gross effects in the sighted group (Hertrich et al., 2006).

It feels normal because it is a learned ability; to continue with the example of these blind folks - they haven't been able to read for a long time and the only way they communicate through written materials is via screen readers. The spoken word is slower than reading so they crank up the speeds whenever possible, so they are slowly increasing their speech understanding skills.

<sub>References
- Dietrich et al., Front Hum Neurosci (2013); 7: 701
- Hertrich et al. Brain and Language (2006); 124(1): 9-21</sub>

Answer 2

1. There is probably not a large difference from what occurs during normal listening--and that is likely why speed listening is effective:

   a. The reason it "feels" normal is the same that any other sensory stimuli feel normal after a while: your brain habituates to the patterns of your sensory experience. If you increased the pitch of the podcast without changing the speed, for example, eventually the chipmunk voices would sound normal.

   b. It is almost certainly a learned ability, or, really, just a different comfort zone. The average rate of self-talk (i.e. verbal thoughts) is usually given as around 400 words-per-minute, and this is also the upper end of normal reading rates (200-400 WPM). According to the National Center for Voice and Speech, "The average rate of speech for English speakers in the United States is about 150 words per minute". Narrators for audiobooks and podcasts often speak somewhat slower than this to compensate for the lack of visual cues. So, at 3X speed, the typical audio track approaches the upper limits of typical verbal processing (by the way, reading, thinking, and listening ALL activate the same parts of auditory cortex associated with language). So, unlike with speed reading, it is entirely reasonable that a majority of people could learn to comprehend audio content at 3X the typical rate (but see caveat below). However, speed listening is going to require more attention. People who complain they cannot understand at higher listening rates likely should say they cannot do the simultaneous commentary/reflecting/daydreaming that they normally do simultaneously.

2. It depends on the content, but probably not for most podcasts. Let's assume that the baseline rate of the track is such that, at 3X, it doesn't go above 400 WPM. Most podcasts take care to provide narratives that are well-structured and easy to follow--that's why they're enjoyable to listen to. All the verbal reasoning necessary to comprehend the content is included in the track. But now imagine listening to a math or physics textbook at that speed. In this case, while sufficient information is provided for understanding, it requires quite a bit of cognitive "unpacking" (matching with context, relating concepts, generating cognitive models, etc.), a bit like a zip file. I remember my math professor saying even trained mathematicians can spend 30 min understanding a single page! Another caveat is memory. Even though you may comprehend every word of the track, you definitely won't remember it in every detail. And remember that commentary/reflecting/relating you gave up in focusing your attention? Studies show that's actually important to learning and memory (not so with daydreaming, of course). But as long as you pause when you come across an interesting point to really let it sink in, you'd probably solve that problem as well.