Phonological Illusions

Why do we sometimes hear things that aren't there? When does the way we process language leave cracks for illusions to appear? In this week's episode, we talk about phonological illusions: what varieties there are, the processing strategies that lead to them, and how sometimes, we even try interpreting sounds as language when they aren't linguistic at all.
If you want to see the full video of Peter Ablinger's talking piano, it's over here: • Speaking Piano - Now w...
And if you want a couple of other McGurk effect videos, you can try these two: • The McGurk effect and • McGurk Effect (with ex...
This is Topic #38!
This week's tag language: Dutch!
Find us on all the social media worlds:
Tumblr: / thelingspace
Twitter: / thelingspace
Facebook: / thelingspace
And at our website, www.thelingspa... !
Our website also has extra content about this week's topic at www.thelingspace.com/episode-38/
We also have forums to discuss this episode, and linguistics more generally.
Sources:
Clarke, C. M. & M. F. Garrett (2004). Rapid adaptation to foreign-accented speech. Journal of the Acoustical Society of America 116 (6), 3647-58.
Fernandez, E. M. and H. S. Cairns. (2011). Fundamentals of Psycholinguistics. Wiley-Blackwell.
McGurk, H. & J. MacDonald. (1978) Hearing lips and seeing voices. Nature 264: 746-8.
Warren, R. M. (1970). Perceptual restoration of missing speech sounds. Science 176: 392-3.
Looking forward to next week!

[@Yackielot]

I can't understand how underrated these are. On par with a crash course series. This was amazing, very trippy. Like even when I KNEW what you were doing, if I repeated the video, my mind /still/ insisted on hearing 'splash' or 'da'. It reminds me of the moment we all realised that context changes how we interpret colour (with the whole white/gold blue/black dress fiasco).

[@thelingspace]

Yackielot Thanks! We hope they keep spreading, too. If you like the videos, tell your friends about them! And yeah, the thing that makes these such strong illusions is that basically, your brain just uses these processing routes automatically, and there's nothing to be done about them consciously on your end. It's the price of getting to do language quickly. And you're right about the color perception stuff being similar. Sometimes our system gets hit just the right way, and all our regular tricks just push us the wrong way. ^_^

[@CorneliusSneedley]

Thanks, Moti, very interesting video. I was formerly blissfully unaware of this; I find it both fascinating and frightening that our senses can so easily deceive us. The talking piano is amazing, everyone should have a look at that video.

[@thelingspace]

Cornelius Sneed Yeah, that's a really cool phenomenon. I really like that piano. And sometimes our brains don't quite settle things right! It's worth it for what those processes allow us to do, but yeah. Sometimes things just go awry. Thanks for the comment! ^_^

[@YumiSparkles]

New subscriber here, I've been going through these videos chronologically all night and been loving them so far! As for this video, I want to here is what I found for myself with the two little demonstrations you did: I actually heard a sort of in-between of 'sklash' and 'splash' when you said it, possibly more towards the 'sklash' side of things, but quickly interpreted it as 'splash', even if I thought that wasn't quite what you said. For the demonstration of the McGurk effect in this video, I don't hear 'ga', 'ba', or even 'da'. What I hear is 'la', both while watching and while only listening, and I can't make myself hear anything else.

[@felipevasconcelos6736]

I noticed the ezample. Mostly because I was using subtitles (love your channel because of them).

[@thelingspace]

+Felipe Vasconcelos Yeah, they can be a bit of a giveaway sometimes, but I think they're worth it overall. ^_^

[@felipevasconcelos6736]

The Ling Space They are indeed. As I'm not a native speaker, there are always some expressions that I never heard before, so I always keep learning!

[@alinafdima]

I actually heard all 3 sounds in the McGurk part! the first two were weird da and ga, and the last one is closer to a ba (also a bit weird). I was so confused that I couldn't decide between either of them, even though I knew about the effect before. And it should be noted that I am not a native speaker.

[@jordanschriver4228]

4:20 I actually heard "slash", not "splash". 7:37 I actually heard the voiced uvular implosive. I might process speech differently because I'm autistic.

[@kira7573]

I love your channel! These videos are so good I'm surprised they have so few views!!!!

[@andyl666]

i also heard 'slash'. maybe because with the sound taken out it sounds close to 'slash' with a partially voiceless l, as l is often pronounced after another consonant (more in British English than American I believe?). I wonder if British English speakers would be more likely to hear 'slash' instead of 'splash'?

[@Glooberloob]

i heard sklash :o i was so confused when he said it i just found this channel when i was looking up phonemic restoration, and im glad i did

[@ReadHeadPat]

I met one of you guys at VidCon this year (at the BrainScoop Q and A i think). Just subscribed because this was a really cool video! Glad I found your channel :D

[@thelingspace]

ReadHeadPat Good to have met you! That'd have been our director, Adele. Thanks for the kind words about the video, and we hope you like sticking around. ^_^

[@Yotanido]

I don't actually hear "da", but I don't hear "ba", either. Even when I know that that is actually being said. That's bizarre. I actually head a really weird "ga". I can tell something is off, but it does sound like "ga" with the visuals...

[@thelingspace]

+Yndostrui Interesting! How do you feel about the other McGurk effect videos we posted in the video description: kzbin.info/www/bejne/l3ezpZZud7uXgM0 and kzbin.info/www/bejne/oKXWl5yah9Zlqtk ?

[@Yotanido]

+The Ling Space In the first one it's the exact same thing. In the second one, I first actually heard 'da', but then it went to the weird 'ga'. I can like mentally switch between the two. (Like you can with some optical illusions - best way I can describe it :D) Now that I discovered this, I can actually make myself hear 'da' in the first one, too. Never 'ba', though. Never 'ba'...

[@ThePolyglotGrind]

I wasn't sure where to ask this, but anyway. I'm curious. As we all know there's a lot of immigration around the world and I was wondering if for example a large group of chinese or Punjabi speakers moved to Arizona let's say, and they integrated into society, but still spent significant amount of time within their respective linguistic groups speaking their respective native languages and this continued through generations (given the fact the language survived the generations) would the form of Punjabi or chinese spoken in Arizona differ greatly from the form spoken in their respective countries of origin? Pretty much what I'm asking is how languages evolve when a large group of people form a community outside of the language's country of origin.

[@thelingspace]

The Polyglot Grind Thanks for the question! I think it's pretty safe to say that they would evolve away from the base language, yeah. Even without the geographical divide and the people residing within a different group of speakers, it's hard not to expect that they'd change, given that these changes happen even within the same country. Look at the different varieties of English in the UK (or the US, or Canada, etc.), or French within France (let alone Canada, various African countries), or Korean between North and South Korea. But if you put an immigrant group somewhere new, chances are they'll adopt a lot of words from the region they've moved to, which'd be one difference. But even if they don't, the independence from the original language and the community building would lead to divergences from the original language. We can see this pretty easily just in, say, the different varieties of Spanish spoken from different places the Spanish colonized. Even if the original groups were different, and even if they tried to maintain their Spanish from the native populations, there's still a lot of changes that happened. Hope this answers your question! ^_^

[@davidphilipsmusic]

Heard da first time, closed eyes and heard ba. Crazy stuff. Your videos are great, many thanks :)

[@laurenttits459]

Cool illusions!

[@MrL1193]

It's weird, but when you were demonstrating the McGurk effect, I actually did hear "ba" right off the bat. Maybe it's because your enunciation was too clear or something, but I just can't seem to hear "da" or "ga," even when I'm watching your mouth. I know I've seen the McGurk effect demonstrated elsewhere before (not that I remembered its name before you explained it in the video), and I remember that that time, it sort of worked in that I couldn't quite figure out what sound I was hearing, but I don't think it ever actually tricked my brain into committing to the wrong sound.

[@thelingspace]

MrL1193 That's interesting! If you cover your ears, or mute the audio, what happens? Can you tell then that I'm saying [ga]? I'm definitely curious here! There are a couple of links to other demonstrations of the McGurk effect in the description of the video; neither of those work for you, either? It doesn't really sound like a natural [da] to me, more of a weird one, but I'd still say I'm hearing a [da].

[@MrL1193]

The Ling Space Hm, I can definitely see the "ga" when I watch you with the sound muted. In the first of the linked videos (the 5 second one), I still hear "ba," but in the second one, I can indeed hear it as "da" (albeit a somewhat strange one). I feel like in that video, the "ba" is pronounced in more of a droning voice compared to your demonstration and that of the other linked video. Also, now that I think about it, I remember that the first time I saw the McGurk effect demonstrated (the time that it confused me), it was also pronounced in a fairly droning fashion. I wonder if that might be helping to mask the "b" sound?

[@thelingspace]

MrL1193 Maybe. The droning flat nature may be getting rid of more of the prosodic information that is allowing you to get around the effect, possibly. The nice thing about the McGurk Effect is that it's been demonstrated in a lot of places, so if you look around more, you might be able to check to see if this makes a difference for you. I don't know if anyone has studied that, but it could be interesting to look at whether the flatness of the presentation makes a difference. ^_^

[@ellisivtorvik4713]

Can you please do a video on generative phonology? Sincerely, struggling student

[@thelingspace]

+Ellisiv Ørnes Vestre You're right, we haven't done a video on phonology stuff instead of more phonetic stuff for a long time. We'll try to get out one on that some time early next year. ^_^

[@dragoncurveenthusiast]

I have the suspicion that we only have/use bottom up processing when we are in the process of learning a language. We are not that flexible replacing unclear words and sounds because we don't know yet that the word didn't make sense. It could have been a word that we just don't know yet. I think the same goes for reading handwriting. If you know the language well, you are better at guessing what the word could be, because you can compare it to all the words it could mean. Those are just assumptions from my own experiences learning new languages, but I'd be curious if someone has looked into this scientifically!

[@thelingspace]

I think I'd say that even for native speakers of a language, we kind of need to be doing both at the same time. Part of the challenge of language learning is not having access to good top-down strategies to guide us and having to rely more heavily just on bottom-up processing, it's true. In a language we're proficient in, we can rely a lot on expectations and probabilities from top-down processing. But in order to deal with novel or weird sentences, or with cases where interpretation or grammaticality hinges on hearing and processing just one sound, we need to rely on bottom-up processing, too. If both aren't working at the same time, we're going to have a hard time of it. That'd be what I'd say about this, I think. ^_^

[@dragoncurveenthusiast]

The Ling Space Thanks for the reply! Good to know what somebody from the field thinks about my empirical hypotheses I pieced together :-)

[@JqlGirl]

For the first few months I was learning Finnish, I would always hear the /äy/ diphthong with a phantom L sound in the middle of it.

[@thelingspace]

+JqlGirl I can picture that, actually, listening to it. The acoustics of it in transit might be pretty close to an [l]. I'll try to find some spectrograms or formant values and see if that's really true, but just picturing it, it could be. ^_^

[@JqlGirl]

+The Ling Space Specifically, I always heard the /l/ when the /äy/ preceded /t/, such as in "näyttää"

[@thelingspace]

+JqlGirl That makes sense, too! Again, from a coarticulation perspective - your tongue starts moving towards the next sound while you're still making the previous one. And that would make that be fairly [l]-ish, I bet. ^_^

[@TalhaBedir]

so interesting

[@t.k.abrams4720]

I definitely heard /b/ but when you said I should have heard /d/ I went back and listened closely and watched closely and kind of heard it as /ɖ/

[@laripolimata]

I heard a weird "slash", maybe because I speak English as a foreign language and the word was pronounced out of a context

[@jurryaany]

Hmm this channel could use quicker cuts to kill the noiseless gaps in between, also less yelling.

[@teacherdkennedy]

I went to Thailand recently and they were speaking Chinese on the airplane back. I don't know any Chinese, but after a but I started to hear random English words in the announcements, like "Chinese Chinese Chinese Barbie Chinese Chinese. ". It was really weird. Of course I had also been awake a long time by then....

[@RedHair651]

I clearly heard /ba/ even when watching.

[@chajadan]

I was hearing a G sound, like the G version of this sound: en.wikipedia.org/wiki/Voiced_alveolar_lateral_fricative But further back, more like this one: en.wikipedia.org/wiki/Voiced_velar_fricative And that talking piano is crazy because, it's just a piano -- there's no tongue! No places of articulation!

[@thelingspace]

Charles J. Daniels Yeah! The piano is really cool. We'll talk about it in the future, but speech does have characteristic frequencies for the different sounds, and the piano does play on that. It still doesn't sound natural, but the fact that we can make it out is really cool. And for the sounds, that's really interesting! The sound was definitely a [ba], and it shouldn't be that far back. Sometimes the way our brains work these dissonances out isn't 100% predictable - I've heard about people hear [wa] and [la], too. Thanks for the comment! ^_^

[@chajadan]

The Ling Space Yeah I definitely heard the ba clearly when I shut my eyes. I think I didn't hear da because my eyes didn't see an appropriate tongue -- the mouth canal seems so open.

[@thelingspace]

Charles J. Daniels That's interesting! I wonder, there's a couple of other McGurk effect video links in the video description - what happens to you when you watch those? Do you hear the same sound for those, as well?

[@chajadan]

The Ling Space The shorter one also sounds like a ga to me. If I listen a few times, I can hear some da's, but the first first one always sounds like ga to me. The longer video comes out much more a da, but notice the tongue rise is way more prominent. And with my eyes closes, it much less obviously a ba than yours or the shorter referenced McGurk. It almost sounds pa'y to me.

[@thelingspace]

Charles J. Daniels Interesting! Thanks for letting me know. I'll be putting a Tumblr post about this soon, too. ^_^

[@eratonysiad2582]

Nice, I heard "slash".

[@thelingspace]

Chris vd. Heijden Interesting! It's cool to hear what's working for people and what's not. I checked all of them on a few friends before I put them up to be sure they worked, but it's cool. Not everyone works the same! ^_^

[@naouri86]

I also heard "slash". Perhaps because 'slash' is a word, there is no confusion in language processing areas of the brain, and therefore no attempted phoneme restoration... maybe?

[@piouppioup]

Wow, the McGurk effect is so weird! Does that mean we all low-key lip-read?

[@MK-je7kz]

I had my friend watch a video about McGurk effect, but she completely ignored the visual side of it. I assume it was because she is French, who likes foreign movies and so is used to dubbing and to ignore the visual mismatch.

[@linguaphilly]

I heard Ba three times :(

[@thelingspace]

+jortjuuuuuh Hmmm. Well, this effect works better on some people than others. Have you tried the other videos that we linked to in the video description? Some people who've had more trouble with ours have found it easier to get those (and vice versa).

[@linguaphilly]

+The Ling Space I tried the others just now, I kept hearing ba there too. I hope I don't have some kind of brain disformity lol! Maybe it's because of my background in music and extensive training in discerning between different sounds?

[@thelingspace]

+jortjuuuuuh That may well be it, yeah. If you really are used to keeping your mind away from the visuals and just honing on the specifics of what the sound is, I could definitely see that making the effect harder to achieve.

[@sugarfrosted2005]

I heard stlash. I have no idea why.

[@Glooberloob]

i heard sklash :o i was so confused when he said it

[@aerobolt256]

Strangely I heard something along the lines of /skʔplæʃ/ and /gbɑ/