Phonaesthetics, or “The Phrenology of Language”

 

Stop me if you’ve heard this before: French is a beautiful, romantic language; Italian sounds like music; Spanish is passionate and primal; Japanese is aggressive; Polish is melancholic; and German is a guttural, ugly, unpronounceable mess (Ha! Tricked you! You couldn’t stop me because I’ve written all of this down way before now. Your cries and frantic gesticulations were for naught.)

We’ve all heard these judgements (and many others) repeated multiple times over the course of our lives; not only in idle conversation, but also in movies, books, comics, and other popular media. There’s even a series of memes dedicated to mocking how German sounds in relation to other European languages:

“Ich liebe dich” is a perfectly fine and non-threatening way of expressing affection towards another human being

What you might not know is that this phenomenon has a technical name in linguistics: phonaesthetics.[1]

Phonaesthetics, in short, is the hypothesis that languages are objectively more or less beautiful or pleasant depending on various parameters, such as vowel to consonant ratio, presence or absence of certain sounds etc., and, not to put too fine a point on it, it’s a gigantic mountain of male bovine excrement.

Pictured: phonaesthetics

Let me explain why:

A bit of history

Like so many other terrible ideas, phonaesthetics goes way back in human history. In fact, it may have been with us since the very beginning.

The ancient Greeks, for example, deemed their language the most perfect and beautiful and thought all other languages ugly and ungainly. To them, these foreign languages all sounded like strings of unpleasant sounds: a mocking imitation of how they sounded to the Greeks, “barbarbarbar”, is where we got our word “barbarian” from.

In the raging (…ly racist) 19th century, phonaesthetics took off as a way to justify the rampant prejudice white Europeans had against all ethnicities different from their own.

The European elite of the time arbitrarily decided that Latin was the most beautiful language that ever existed, and that the aesthetics of all languages would be measured against it. That’s why Romance languages such as Italian or French, which descended from Latin[2], are still considered particularly beautiful.

Thanks to this convenient measuring stick, European languages were painted as euphonious ( ‘pleasant sounding’), splendid monuments of linguistic accomplishment, while extra-European languages were invariably described as cacophonous (‘unpleasant sounding’), barely understandable masses of noise. This period is when the common prejudice that Arabic is a harsh and unpleasant language arose, a prejudice that is easily dispelled once you hear a mu’adhin chant passages from the Qur’an from the top of a minaret.

Another tool in the racist’s toolbox, very similar to phonaesthetics, and invented right around the turn of the 19th century, was phrenology, or racial-biology, the pseudoscience which alleged to be able to discern a person’s intelligence and personality from the shape of their head. To the surprise of no one, intelligence, grace and other positive characteristics were all associated with the typical form of a European white male skull, while all other shapes indicated shortcomings in various neurological functions. What a pleasant surprise that must have been for the European white male inventors of this technique![3] Phrenology was eventually abandoned and widely condemned, but phonaesthetics, unfortunately, wasn’t, and it’s amazingly prevalent even today.

To see how prevalent this century-old model of linguistic beauty is in popular culture, a very good example are Tolkien’s invented languages. For all their amazing virtues, Tolkien’s novels are not exactly known for featuring particularly nuanced moral actors: the good guys might have some (usually redeemable) flaws, but the bad guys are just bad, period.

Here’s a brief passage in Quenya, the noblest of all Elven languages:

Ai! Laurië lantar lassi súrinen,

Yéni únótimë ve rámar aldaron!

Yéni ve lintë yuldar avánier

Mi oromardi lissë-miruvóreva

[…]

Notice the high vowel-to-consonant ratio, the prevalence of liquid (“l”, “r”), fricative (“s”, “v”) and nasal (“n”, “m”) sounds, all characteristic of Latinate languages.

Now, here’s a passage in the language of the Orcs:

Ash nazg durbatulûk, ash nazg gimbatul

Ash nazg thrakatulûk, agh burzum-ishi krimpatul

See any differences? The vowel-to-consonant ratio is almost reversed, and most syllables end with a consonant. Also, notice the rather un-Latinate consonant combinations (“zg”, “thr”), and the predominance of stops (“d”, “g”, “b”, “k”). It is likely that you never thought about what makes Elvish so “beautiful” and “melodious”, and Orcish (or Klingon, for that matter), so harsh and unpleasant: these prejudices are so deeply ingrained that we don’t even notice they’re present.

So why is phonaesthetics “wrong”?

Well, the reason is actually very simple: beauty is subjective and cannot be scientifically defined. As they say, beauty is in the eye of the beholder.

Not this beholder.
Image copyright: Wizards of the Coast

What one finds “beautiful” is subject to change both in space and in time. If you think German’s relatively low vowel-to-consonant ratio is “harsh”, then you have yet to meet Nuxálk.

Welcome to phonaesthetic hell.

Speaking of German, it is actually a very good example of how these supposedly “objective” and “common sense” criteria of phonetic beauty can change with time, sometimes even abruptly. You see, in the 19th century, German was considered a very beautiful language, on par with Italian or French. A wealth of amazing prose and poetry was written in it: it was probably the main language of Romantic literature. It was also the second language of opera, after Italian, and was routinely described as melodious, elegant and logical.

Then the Nazis came.

Nazis: always ruining everything.

Suddenly, Germans were the bad guys. No longer the pillars of European intellectual culture, their language became painted as harsh, aggressive, unfriendly and cold, and suddenly every Hollywood villain and mad scientist acquired a German accent.

So, what’s the takeaway from this long and rambling rant?

No language is more, or less, beautiful than any other language. All languages have literature, poetry, song and various other ways to beautifully use their sounds for artistic purposes, and the idea that some are better at this than others is a relic from a prejudiced era better left behind. So next time you feel tempted to mock German for how harsh and unpleasant it sounds, stop and think that maybe this is not actually what you think, and that you’ve been programmed by a century of social prejudice into thinking so.

And read some Goethe, you’ll like it.

Stay tuned for next week, when the amazing Rebekah will bring you on the third leg of our lightning trip through Phonphon!

  1. Phonaesthetics also has a different meaning, which is the study of how certain combinations of sounds evoke specific meanings in a given language. Although this form of phonaesthetics has its problems, too, it is not what I’m talking about in this post, so keep that in mind as we go forward.
  2. See our post on language families here.
  3. First the men assumed that the female skull was smaller than the male, and this was obviously a sign of their inferior intelligence. Later, however, they found that the female skull was larger, so they came up with the idea that this meant females were closer to children, and thus the male was still more intelligent! – Lisa

Phonology 101: Consonants, or “Let’s Make a Sandwich”

Well, well. Here we are again. Back for more Phonology 101. Today, we’re finally going to start putting names to faces, or, uh, sounds. That’s right, it’s time to talk about consonants!

Phones (as we learned to call speech sounds last time—catch up here) can be broadly divided into consonants and vowels.1 As you probably know, words can be divided up into syllables. A quick and dirty way to separate vowels and consonants is that vowels form the nucleus, i.e. the middle, of a syllable, and consonants are the sounds that go around the outside.2 If syllables were sandwiches, consonants would be the bread and condiments and the pickles and the lettuce and all that good stuff you put around the meat, i.e. the vowels.

But really, the main distinguishing characteristic of a consonant is that the flow of air through the vocal tract is impeded in some way. The nature of the impediment is called the manner of articulation.3 Together with place of articulation and voicing (which we discussed in part one of Phono 101—here), the manner of articulation is one of the three major defining characteristics of a consonant.

If you’ve been studying up on your IPA, the manner of articulation is listed in the left column of the pulmonic consonants chart:

IPA Chart, http://www.internationalphoneticassociation.org/content/ipa-chart, available under a Creative Commons Attribution-Sharealike 3.0 Unported License. Copyright © 2015 International Phonetic Association.

Today, I’m going to go over several types of consonants, which will cover the bulk of the chart. The consonants I won’t cover are trills and taps (which you can see on the chart above), and clicks, implosives, and ejectives (which are listed elsewhere on the complete IPA chart, viewable here). These are much rarer in the world’s languages and, wouldn’t you know it, a bit trickier to grapple with. The sounds I will cover are all present in English, and that’s where I’ll be pulling my examples from. This is not because English is somehow more exemplary than other languages, but I’m writing in English, so I assume you, my trusty reader, have at least some familiarity with its sounds.

Our first type of consonant is called a plosive or a stop (the terms are interchangeable). (Ideally, I would say stops are wheat bread or mayonnaise or some other sandwich-making component, but this is a shaky metaphor that I’m afraid will quickly fall apart if spread too far.4) In these sounds, the airflow is completely blocked somewhere along the vocal tract and then released. The sound is stopped, then explodes outwards. (Most of the consonants have helpful names like that. Thankfully, we linguists don’t always use fancy Latin terms for things.) For example, when producing [b], both lips come together and part, stopping and releasing the flow of air. When producing [t], the closure is made by placing the tongue along the alveolar ridge. (I asked you before to try saying different sounds to get a feel for the places of articulation. You should definitely try that again today, this time paying attention to the manner of articulation. Go on. You know you want to.)

Another type of consonant is the fricative. Here, the flow of air is highly restricted but not completely cut off. These sounds tend to buzz or hiss, and the name of the game here is friction. Excellent examples are [v] (a voiced labiodental fricative) and [s] (a voiceless alveolar fricative). Fricatives are fun because you can drag them out as long as you want as long as you’ve got enough breath support.

Now sometimes, occasionally, when a stop and a fricative really love each other, those two crazy kids get together and form an affricate. An affricate is a stop followed in such close sequence by a fricative, they overlap to create what sounds like one phone. English has a couple of these. The first and last sounds in church (/t͡ʃ/) and judge (/d͡ʒ/) are affricates. There’s some debate over whether affricates should be treated as one sound or as a sequence of two.

Stops, fricatives, and affricates can, generally, be either voiced or voiceless. Due to how they’re produced, this isn’t necessarily true for some of the other types of consonants.

Some of those perpetually voiced sound are known as approximants. In these sounds, the airway is restricted, but the air passes around the sides of the tongue or passes through the narrowed areas without the friction characteristic of fricatives. This category includes sounds like [l], [ɹ], and [w]. “Approximants” is actually kind of a catch-all term for some of the most complicated consonants to produce, and they can be broken up into several subcategories. We’re not going to dive into those right now; I don’t want to make your head hurt more than I have to.

In all of these sounds (stops, fricatives, affricates, and approximants), the air passes through the oral cavity. For some sounds, the velum is lowered, allowing the sound to pass through the nasal cavity, like in [m] and [n]. This one’s easy: these sounds are called nasals.

That’s the whole grocery list, so let’s get back to our tenuous sandwich analogy. Some sandwiches/syllables are simple—just two pieces of bread and a slice of roast beef, like cat. Some have mayonnaise and tomatoes and cheese and all sorts of other fixings, like strength (the linguistic term is consonant clusters). Some sandwiches/syllables are open-faced, like do. No matter how you serve it up, put all the ingredients together and you wind up with something delicious.

In two weeks, I’ll be back to finish off Phonology 101 with an exploration of that phonemic delicatessen, vowels. Next week, Riccardo will be here to bend your mind on the subject of linguistic prejudice and a little something called phonaesthetics.

Notes

1Finally! Something intuitive, am I right?
2There are a couple consonants that occasionally blur this line, but contrary to what you might expect, none of them are represented by the letter Y.
3Technically, vowels also have a manner of articulation, but it’s the same for all vowels. Manner of articulation isn’t really important for vowels, but it’s crucial for consonants.
4Get it? “Spread”? I’m sorry.

The Sapir-Whorf Hypothesis

 

“the Sapir-Whorf hypothesis is the theory that the language you speak determines how you think”

 

So says the fictive linguist Louise Banks (ably played by Amy Adams) in the sci-fi flick ‘Arrival’ (2016). The movie’s plot relies rather heavily on the Sapir-Whorf hypothesis, also known as the principle of linguistic relativity, so heavily in fact that the entire plot would be undone without it.

But what is the Sapir-Whorf hypothesis, really? Before digging into why ‘Arrival’ may have gotten it a bit… well, off, a word of caution: If you haven’t seen the movie (and intend to do so), go ahead and do that before reading the rest of this post because there will be SPOILERS!!!


Now that you have been duly warned, let’s get going.

The Sapir-Whorf hypothesis is, in a way, what Louise Banks describes: it is in part a hypothesis claiming that language determines the way you think. This idea is called linguistic determinism and is actually only one half of the Sapir-Whorf hypothesis.

Commonly known as the “strong” version of Sapir-Whorf, linguistic determinism holds that language limits and determines cognitive categories, thereby limiting our worldview to that which can be described in the words of whatever language we speak. Our worldview, and our way of thinking, is thus determined by our language.

That sounds pretty technical, so let’s use the example provided by ‘Arrival’:

The movie’s plot revolves around aliens coming to earth, speaking a language that is completely unknown to mankind. To try to figure out what they want, the movie linguist is called in. She manages to figure out their language pretty quickly (of course), realising that they think of time in a non-linear way.

This is quite a concept for a human to grasp since our idea of time is very linear. In western societies, we commonly think of time as a timeline going from left to right, as below.


 

Let’s say that we are currently at point C of our timeline. We can probably all agree that, as humans, we cannot go back in time to point A, right? However, in ‘Arrival’, we are given the impression that the reason we can’t do that is because our language doesn’t let us think about time in a non-linear way. That is, because our language doesn’t allow us, we can’t go back in time. Sounds a bit wonky, doesn’t it?

Well, you might be somewhat unsurprised to hear that this “strong” version has been discredited in linguistics for quite some time now and, for most modern-day linguists, it is a bit silly. Yet, we can’t claim that language doesn’t influence our way of thinking, can we?

Consider the many bi/multilinguals who has stated that they feel kinda like a different person when speaking their second language. If you’ve never met one, we bilinguals at the HLC agree that we could vouch for that fact.

Why would they feel that way, if language doesn’t affect our way of thinking? Well, of course, language does affect our way of thinking, it just doesn’t determine it. This is the ‘weak’ version of Sapir-Whorf, also known as linguistic relativism.

The weak version may be somewhat more palatable to you (and us): it holds that language influence our way of thinking but does not determine it. Think about it: if someone were to point out a rainbow to you and you had no word for the color red, you would still be able to perceive that that color was different from the others.

If someone were to discover a brand-new color (somewhat mind-boggling, I know, but just consider that), you would be able to explain that this is a color for which you have no word but you would still be able to see it just fine.

That might be the most clear distinction between linguistic determinism and linguistic relativism: the former would claim that you wouldn’t be able to perceive the color while the latter would say that you’ll see it just fine, you just don’t have a word for it.

So, while ‘Arrival’ was (at least in my opinion) a pleasant waste of time, when it comes to the linguistics of it, I’d just like to say:


(Oh, and on a side note, the name of the hypothesis (i.e. Sapir-Whorf), is actually quite misleading since Sapir and Whorf never did a collaborate effort to formalise the hypothesis)

Tune in for more linguistic stuff next week when the marvellous Rebekah will dive into the phonology of consonants (trust me, you have a treat coming)!

 

Phonology 101: Phonemes, aka phonology’s little superheroes

Welcome back to Phonology 101! Last time, we talked about the vocal tract and two major features of speech sounds, namely place of articulation and voicing (catch up here). Today, we’re going to stop using the word “sound” in favor of some more precise terms: phone, phoneme, and allophone. You may also find it useful to catch up on IPA (here); I’ll be introducing some important notation, and moving forward, I’ll be using IPA more and more.

Individual speech sounds are called phones. In talking about speech production, we keep distinguishing between phonetics (the concrete/physical) and phonology (the abstract/mental). Phones exist on the phonetic plane. Our mental representations of the sounds in our language are called phonemes. If your brain was a computer running a language program, phonemes would be all the sound files. As mentioned previously, the International Phonetic Alphabet is the standard for transcribing phones and phonemes. When using IPA, brackets put around the symbols indicate whether phonetic or phonological data is being represented. Phones go in square brackets (e.g. [a]), and phonemes go between slashes (/a/). On both levels, this notation can be used for individual sounds, complete words, or even phrases and longer utterances.

More technically, a phoneme is “a minimal contrastive sound unit of a language.”1 What that means is, phonemes cannot be broken down into smaller units, and switching a single phoneme for another can make the difference between one word and another. To determine the phonemes of a language, linguists look for something called minimal pairs, i.e. pairs of words that differ in pronunciation by a single phone. An example of a minimal pair is mine /maɪn/ and kine /kaɪn/,2 which demonstrates that /m/ and /k/ are phonemes in English. As you can guess, minimal pairs have a tendency to rhyme, but the differing sounds don’t need to be at the beginning of words. They can also come at the end, as in mine /maɪn/ and might /maɪt/, or in the middle, as in can /kæn/ and kin /kɪn/.3 (These examples also nicely illustrate the point we keep emphasizing about how separate spelling and pronunciation actually are.)

Now, the thing about phonemes is, the sound we think we’re producing isn’t always technically the sound that comes out. Based on where it is in a word or the sounds surrounding it, a single phoneme might be pronounced in several ways. The physical manifestations of a phoneme are called allophones. Some phonemes might have only one allophone, while others might have several. Some allophones might be associated with more than one phoneme. I’ve had several teachers who explained it like this: Sometimes Superman is Superman, and sometimes Superman is Clark Kent. Sometimes, he’s even Kal-El. How he appears depends on the environment he’s in, but at the end of the day, it’s all the same guy. A linguistic example: In my dialect, I pronounce rot as [ɹɑt], but rotten is [ɹɑ.ʔən]. (Like generic phones, allophones go in brackets.) My “t” sound in rotten is something called a glottal stop. Even though it’s a different phone, my brain says the middle sound in rotten is the phoneme /t/ (an alveolar stop). This comes out when I enunciate it and it sounds more like [ɹɑ.tən]. [ʔ] is an allophone of /t/. (As we haven’t really talked about any specific sounds yet, I’ve tried to pick examples that would be kind of intuitive for most speakers. Even so, it was hard to avoid using some sounds and symbols that may be making you go, “Huh?” Here’s an interactive IPA chart with sound clips if you’d like to hear what I’m talking about.)

Some allophones are pretty obvious to spot, but others are more subtle. In ram [ɹæ̃m] the vowel takes on a nasal quality not present in rat [ɹæt] through the influence of the following nasal consonant /m/. This is how allophones tend to come about. In actual speech, no sound is produced in a vacuum. In the interest of ease of articulation, our mouths carry over features from previous sounds or, as in the case of ram, anticipate upcoming sounds. This process is highly predictable and regular. A phoneme will consistently produce the same allophone in a given environment.

Okay, so what? This is all well and good for linguists who just want to Know, but does it have any impact on your life? Well actually, yes. Every language variety has its own phoneme inventory. Phonemes in one language may only be allophones in another (or may be missing entirely). When we pick up a language as children, our boxed set of language features includes a list of sounds we consider phonemes. As we grow up, this internal phoneme inventory solidifies and becomes very hard to alter. This explains why people have an accent when they learn a second language—they’re unconsciously using their native phonemes.

Our phoneme inventories don’t just influence the way we pronounce words. They also filter how we process what we hear. In my dialect, I pronounce merry, marry, and Mary the same. In some dialects here in the United States, these words contain three different vowels and therefore have three distinct pronunciations. I’ve studied phonology long enough that I can carefully reproduce these pronunciations (when I remember which vowel goes with which word), but I still can’t hear the different when a native speaker of this dialect produces these words. For me, all these vowels get lumped into the same phoneme, so that’s how I hear them.

This isn’t a big deal for a situation like mine, where the phoneme difference is slight and it will be easy for me to figure out what’s being said. The differences between phoneme inventories can get more pronounced4, and interfere more with comprehension, when you’re speaking a second language. For example, in Arabic, [b] and [p] are not distinct phonemes, so it’s hard for a native Arabic speaker to distinguish between English barking and parking.

And this is why we’re wading through all this terminology and technical swampland: Language is a real world phenomenon with real world consequences. As in the Arabic example, phonology can have an effect on actual communication, as can morphology, syntax, semantics, and the other basic disciplines of linguistics. Once we’re done laying the groundwork, we can get to the fun stuff (like, what’s up with English spelling? And how do you pronounce GIF?).

There will be two more posts in the Phonology 101 series (for now). In two weeks, we’ll get into the nitty-gritty of consonants, and then in March, we’ll go over vowels. Before we get to that, though, Sabina will be back next week to talk about something called the Sapir-Whorf hypothesis and ask the big questions: Does our language influence how we think?

Notes

1Giegerich, Heinz J. 1992. English phonology: An introduction. Cambridge: Cambridge University Press.
2An archaic plural of cow. I just really wanted to use it.
3Another somewhat archaic word for family and relatives. (But maybe you knew this one.)
4Lol, phonology puns.