Thread by @realscientists

A very good Wednesday, fine folks of twitter!!!! I hope you feel chatty, BECAUSE IT'S LANGUAGE DAY.

As well, Ramadan Mubarak to folks observing Ramadan!!

In our music and evolution talk, we discussed how music may have evolved as a proto-language (and tangents and gifs, which is to say, twitter):

External Tweet loading...
If nothing shows, it may have been deleted
by @realscientists view original on Twitter

Today, we're going to dive in and talk brainssssssss, language, and music.

For this, we're going to start at the commonalities and branch off. It's helpful for me to think of this as a tree. See also, this paper by Stephen Brown: (neuroarts.org/pdf/nyas.pdf)

Let's start at the root. Music and language share a common auditory processing network that responds to sounds, localized the sources, and feeds it to higher-order areas for processing.

Today, I'll get a little more *specific* with brain areas. This processing happens in Brodmann Areas 41 and 42 (temporal lobe subdivisions). Brodmann areas are a standard atlas of cortical structures that are pretty common.
Brodmann:

So sound is processed, and you auditory cortices are all like:

The thalamus (a structure between the upper cortex and midbrain) routes the sound to higher structures in the cortex and then we get into level two: parsing the sound!

PARSING IS OUR WORD OF THE DAY

Here's where the branches diverge. Both music and language contain multiple laaaaayers of information, broadly categorized as syntax, semantics, and emotion

Syntax is, in general, rules; and syntax errors are those that break grammatical rules. Syntax is handled in BA44 (part of Broca's area), but here's where the plot, like a perfectly-made custard, THICKENS.

Lateralization emerges and linguistic syntax is processed on the left whereas musical syntax is processed on the right.

Similar error detection systems are present for both, and you see a distinct EEG pattern when syntax is violated in music and language studies.

If you're interested in this, here are two amazing reviews by Koelsch and colleagues (music):pubman.mpdl.mpg.de/pubman/item/es…
and Friederici (language): physiology.org/doi/pdf/10.115…

This kind of parallel processing in distinct but related brain areas is called a homologue. Neat!

AND WE'RE BACK. Next up is melody! In music, melody is fairly obvious, but in language, it might seem less so. Think about the differences in pitch, stress/inflection, and tone you use when talking.

This is very language-specific, accent-specific, even regional dialect-specific. For example, in English, if you ask a question, you might raise the pitch at the end to signify that a question is being asked.

This is not the case in Finnish. No matter how many years you spend trying to make it so. SARAH.

(sidenote - Finnish is great: )

The musical elements of speech are called prosody. You can get limited information from prosody alone (eg hearing someone through a wall...you might be able to tell if thy're talking/laughing/crying), but paired with the lexical (word-y) elements of speech, it adds nuance.

Think here about "that's great!". Depending on the *prosodic* elements of the speaker, the statement can be extremely enthusiastic, or EXTREMELY sarcastic.

Prosody is processed in the left portion of BA22, back in the temporal area. Neat info here:
umich.edu/~cogneuro/jpg/…

The right portion of BA22 is the music homologue, and processes emotional aspects of melody. Where melody doesn't have lexical elements, it's straight-up emotion rather than prosody.

These are directly homologous, but things get a little fuzzy when we get into semantics. Semantics refer to the meaning of something, and errors can occur without syntactic violations.

Semantic processing in language uses areas in the Inferior Frontal Gyrus, Middle Temporal and Superior Temporal Gyri, and appear to be left-lateralized. Fun fact: a gyrus is a ridge, and its corresponding valley is a sulcus! LEARNING.

Musical semantics are fiddly and treacherous, and a pain in the arse to define. Where meaning is allowed to be so fluid in music, how can you pinpoint errors, or even rules in the first place?

Back to the work of Stefan Koelsch and colleagues:
stefan-koelsch.de/papers/Koelsch…
researchgate.net/profile/Daniel…

Semantics in music uses areas in the Inferior Frontal Gyrus and Superior Temporal Gyrus, just like language does, but isn't as constrained to the left hemisphere as language.

Music:

Now what about lyrics? With music with lyrics, you have EXTRA FUN TIMES because you can have double the syntactic/semantic information to process!

Let's play a game. Is this banger by New Brunswick's finest @mylesdavid ft. @classified a happy song or a sad song?

Winner winner, chicken dinner!

External Tweet loading...
If nothing shows, it may have been deleted
by @jens0331 view original on Twitter

As many of you guessed, the song is both happy and sad. The music and lyrics are incongruous, so depending on what you focus on first, you'll find the song either happy or sad. Or both!

Another *cllllllllassic* example is here:

Despite studying music for a living, I base 100% of my song judgements on language, so it takes me repeated listenings to catch onto the lyrics. Unless the instrumentation is VERY sparse and the lyrics are front-and-centre. Example:

Any more examples, please comment below

So syntax, semantics, and emotion/prosody interact and are bound together to form our understanding of language/music, and as we've seen, they can interact. AND DISAGREE.

But what about when pieces are missing? Here's another musical example. Provided you don't speak Gaelic or Marwari, is this song happy or sad?

(🎶@kaelarowanband)

(it's THE SADDEST. But also, the LOVELIEST)

Now. Not all processing is equal. Music researcher and my actual friend-in-real-life @Anna5ash has studied how musical valence (mood) influences how lyrics are processed, finding that errors are detected more reliably in sad music (Abstract here: journals.sagepub.com/doi/abs/10.117…)

Music's ability to regulate and/or induce emotion has been extensively studied, and if you're interested, go here:
counselorforteens.com/wp-content/upl…
s3.amazonaws.com/academia.edu.d…

That's a bit about music and linguistic syntax and semantics. I'm going to go work on some *stuff*, then we'll continue to work our way up the language/music tree!