So, I’ve been noticing a strong uptick in the use of “the cloud” to refer to online, decentralized storage, computing and program-hosting lately. No shortage of companies are talking about their “cloud computing” services (including my hosting company, Joyent), and it’s become one of those “gotta have it” corporate buzzwords, and it seems like no company’s marketing people will let them release a website, product or service which isn’t in some way cloudy.

This phenomenon itself isn’t noteworthy from a linguistic standpoint (“Web 2.0” seems to have been the same sort of trendy buzzword at some point), but it occurred to me today that for many less-tech-saavy users, this “in the cloud” phrasing might actually be affecting how people view these services, and I think that might be why companies have latched onto this term so strongly.

Let’s take, for example, Apple’s coming “iCloud” information hosting service. Apple is increasingly targeting the non-tech-saavy crowd, and this service, like most of their recent developments, is meant to be largely transparent to the end user. Once you’ve signed up, iCloud will take your music, your photos, your documents, your books, your backups, your contacts, calendars and mail, and any additional information you add in through third party programs, and make it instantly available on all of your devices. As they put it on their own website: “Create a document, iCloud stores it, and pushes it to your devices”. Bam. Magic. You turn the service on and suddenly your data is on all of your devices. Who wouldn’t want that?

A rose by any other name…

They’re doing something linguistically fascinating, though: they make no mention of their machines, servers, databases or storage (at least on the user-facing sites). You create, something cloudy happens, it’s on all your machines. They’ve de-emphasized the middle step. Mind you, Apple’s not the only “cloud” provider to do this (Google Docs de-emphasizes the middle step too), but Apple is certainly the most flagrant. But why bother? Why de-emphasize?

Well, I’ve been toying around with a new hobby. Whenever somebody says “in the cloud”, I’ve found it entertaining to replace it with “on somebody else’s computer”. This simple replacement brings me much joy in the absurdity it creates and how oddly different it makes the act sound:

“Our main working copy of the paper is on somebody else’s computer for group editing, but it’s password protected so nobody but us can edit it”

“My data is safe, I store my address book, mail, passwords, documents and photos on somebody else’s computer.”

“Oh, don’t worry, all of our business information is backed up on somebody else’s computer.”

When put like that, we’re emphasizing the storage, the step that Apple and Google and most of the other cloud providers don’t really want you to think about too much. We’re emphasizing the fact that your data is sitting on a hard drive in another state, watched by a sysadmin who you don’t know. We’re emphasizing that when you put something on the cloud, it’s no longer just yours, and whereas naive users might not hesitate to put something into an amorphous cloud, actually transferring their data onto another computer might tickle enough of their sense of privacy to make them hesitate to upload those bank statements or that racy note from a lover.

In addition, we emphasize the fact that the data is there for the cloud provider to use per the TOS. How much do you think that the recording industry would pay to analyze en masse the music library of hundreds of thousands of iGadget users, even if just for market research? How valuable would it be for a website to figure out where to advertise by asking a company storing passwords “in the cloud” which sites are also visited by people who have stored passwords for their site?

Simply put, putting your data “in the cloud” is amorphous. It’s a mystery, but at the end of it, it just works. Putting your data on somebody else’s computer can get the same ends, but it forces you to think about your data in between your machine and your other devices.

Clouds aren’t necessarily bad

This may sound like a paranoid luddite’s rant, but I use the cloud. I currently use MobileMe, Apple’s current iCloud equivalent, for calendar and address book syncing. I use DropBox to keep my grocery list current across all my devices. I have an SFTP provider for storing backups of my data between at-home backups, and in case of emergency. The cloud can provide, in addition to convenience, a type of security against loss. As a friend of mine pointed out on Google+ (a cloud app):

Somebody else’s computer, with extensive redundancy and backup systems, which makes it much less likely to be lost if my house burns down. It is one kind of security. Not the “no one else will look at it” kind, but the “I won’t lose it in a domestic disaster” kind.

This is certainly true, and one of the best arguments for decentralized, cloud-like computing. Data on my computer in my backpack is fleeting. Data on a well-backed-up server in Dropbox’s massive datacenter is much less likely to be dropped, stolen, lit on fire or broken. These services have a use, whether convenience, ease-of-use for non-tech users, decentralization, or simply as an offsite backup of your data.

The techies who have read this far are doubtless thinking “Come on, I knew this already”. Of course data stored in the cloud is stored on somebody else’s computers. Heck, geeks like myself can likely picture server farms, maybe even imagining the mass storage required. They have a good idea of what sorts of things cloud providers can and can’t do across petabytes of data.

It’s not like I’m blowing the whistle on a massive conspiracy here. Anybody who has thought more than 20 minutes about the idea of a cloud knows that information has to go somewhere, and has deduced that presumably, it’s sitting on somebody else’s computer. Apple’s not choosing to skirt the issue so they can “pull a fast one” on the entire internet, they’re doing it because it’s less intimidating to new users. Google Docs is neglecting to mention their servers because they don’t need to. That’s not why you should be using the phrase “on somebody else’s computer”.

We should be talking about uploading your documents onto somebody else’s computer with grandma when she gets her new laptop and decides that that “iCloud” folder is just like her hard drive. We should be discussing storing information on somebody else’s computer for the clueless CFO who wants to upload the company’s records onto DropBox to be able to work on them from his new iPad.

We should be talking about “the cloud” as storing information on somebody else’s computer so that people will think, if only for a second, about whether they care that that picture, document, or file is something they would be OK with storing on somebody else’s computer.

Because TOSes, “privacy policies”, talking around the issue and other calming language aside, that’s what the cloud is. It’s a vast collection of other people’s computers, and in order to decide intelligently whether you want your data there, you need to know where “there” is.

Tagged with Computers and Software, Corporate Language, Language and Thought, Language Usage, Language, Computers, and the Internet, Words, Phrases, and Idioms | 6 Comments


I’ve been incredibly busy of late with teaching, but I’ve got a pinboard.in account and have begun adding new linguistics-related bookmarks there, as both a resource for my students and for my own enjoyment.

If you’re interested, check out:

My Linguistics Bookmarks

My Phonetics/Phonology Bookmarks

My Natural Language Processing Bookmarks

Thanks, and I hope you all are enjoying your linguistic lives!

Tagged with Computational Linguistics, Conventional Linguistics, Linguistic Mysticism, Notes, Phonetics and Phonology, Recommended Links, Site News | 1 Comment


So, I’ve been teaching an undergrad course on Phonetics and Phonology (hence my recent silence), and have been enjoying the constant search for awkward analogies to make the subject matter a bit more accessible. I find this is especially necessary for phonology, whose actions and motivations can be quite opaque to somebody just approaching the field.

Phonology, just to recap, is the study of how sounds pattern and interact within words and sentences in different languages. Phonology is language specific, in that every language has its own phonology, each having different sounds, and different preferences. This is why Russian allows the /x/ sound (as in Bach, Khrushchev, Chanukah, and Loch Ness) but English doesn’t. This is also why Japanese speakers, when borrowing the word “christmas”, will usually change it into something more like “kurisumasu”, because they simply don’t like having a syllable end with a consonant.

These language-specific preferences are often enforced with phonological rules, which, when an undesirable combination is made, will change one or more of the sounds in the word to make it more acceptable. So, for instance, the plural of “dish” /dɪʃ/ isn’t “dishs” */dɪʃs/, because that would make us put two strident fricatives (a special class of sound containing the sh sound (/ʃ/), /s/, /z/ and the “zh” sound in pleasure /ʒ/) together, and that’s just awkward for us as English speakers. So, a rule kicks in to add in a vowel /dɪʃs/-> /dɪʃəs/, and a second rule kicks in to prevent a voiced/voiceless pairing (Your larynx vibrates during the added vowel, but does not during the /s/) and changes the /s/ to a /z/, its voiced equivalent.

This is somewhat of a simplification, and there’s actually a lot of good evidence that /z/ is the plural (as in “dogs”) rather than /s/, but still, you can see how here, rules are interacting, and changing undesirable combinations into more desirable ones.

For my next class, I had to prepare a slideshow discussing common phonological processes (among other things), and I was having trouble finding a good hook, a good analogy to help the students understand the dynamics at play. So, I asked myself…

Where on earth do petty differences and subtle preferences constantly and violently change combinations according to a set of unpublished, unconscious laws?

Then it hit me: High School.

All the cool phonemes are doing it…

High school, for those of you unfamiliar, is a type of secondary education institution, attended by students in grades 8-12 (usually ages 13-19). Here, there’s no shortage of raging hormones, crazy dramas, and friendships form and dissolve capriciously, and as such, well, it rather nicely parallels the complex interrelationships between sounds.

So, with this analogy in hand, I set out to explain the four most common types of phonological rules/changes in Phonology using High School drama.

#1: Assimilation/Peer Pressure

Assimilation, in phonological terms, is where sounds change to become more similar to one another.

So, for instance, a nasal sound which is usually made with the tongue in the front of the mouth, at the alveolar ridge (like the “n” in “thin”) might move to the velum (at the back of the mouth) when it’s near a velar consonant. To test this for yourself, say the word “thin” a few times, noting the position of the tongue during the /n/, then say “thin kids” repeatedly, and note how your tongue is likely now positioned in the back of your mouth when making that /n/. In phonological terms, the /n/ has become an /ŋ/ before another velar sound. This is an example of “nasal place assimilation”, and is exceptionally common throughout the world.



Assimilation Complete.

In High School, assimilation is an incredibly common process, and is usually referred to as “peer pressure” or “trying to fit in”. If all the other kids in a social group smoke, chances are, new additions will start smoking too. Similarly, if all the sounds around a given segment are voiced, there’s a strong pressure for that segment to become voiced as well.

This analogy is really useful for phonological analysis, believe it or not. If a parent can’t figure out why her child is suddenly snorting lines of Vitamin C, chances are, if the parent examines the friends surrounding her child, she’ll find that they all are doing it too. Similarly, if you can’t figure out why this sound is being nasalized here, just look at its friends. Chances are, they’re all hanging out behind the gym and giving nasality a try.

#2: Dissimilation/Rebellion

Dissimilation is when sounds change to become LESS similar to one another, usually to heighten a contrast which otherwise might not be apparent.

A good example of this in English is the words “surprise” and “berserk”. Surprise, at least nominally, has two r-sounds, one before the /p/ and one after it. But in practice, and especially in fast speech, speakers don’t like that much rhoticity (r-like-ness) in a single word, so they’ll turn that first r-sound back into a schwa (meaning that /səɹpɹajz/ is usually realized as /səpɹajz/). The same goes with “berserk”. When was the last time you heard somebody make both r sounds in berserk?

Marilyn Manson
You’re dissimilar. We get it. Calm down.

Dissimilation is, in my mind, the phonological process that creates rebellion in all its forms. If Mommy and Daddy are well-heeled and conservative, little Jimmy’s gonna be a Maoist. If all the other kids at the school are preppy and straight-laced, you know that Goth kids will emerge.

Just like in phonology, people just can’t stand too little contrast, so they will sometimes take on features only to differentiate themselves from the crowd. If your sound is suddenly velar when surrounded with alveolar sounds, chances are, it’s rebelling and wanting to be different. And sometimes, that desire alone is enough to make a sound change its way of life.

#3: Epenthesis/Showing up on other people’s dates

Epenthesis is a generic term for whenever a sound seems to pop up out of nowhere to help block an awkward or undesirable combination or situation. Often, it’s vowels which are epenthesizing, but consonants pop up too on occasion.

We’ve actually already discussed an example of this, where an epenthesis rule kicks in to add in a vowel when making the plural of “dish” or “wish” or “diss” (/dɪʃs/-> [dɪʃəz]), which serves to prevent awkward /sz/ and /ʃz/ combinations. Similarly, when Japanese speakers are adding in extra vowels when borrowing words like “Christmas” (which turns into “kurisumasu”), they’re doing it to prevent ending any syllables with consonants, which Japanese speakers just do not like doing.



ø -> me / [you]__[that jerk you're dating] (Image Credit)

Epenthesis is a bit trickier, but imagine this scenario. So, Claire like totally likes Daniel. And he’s like, totally cute, and OMG they are TOTALLY meant to be. But then Daniel starts dating Kate. OMG!!

So, rather than just letting the two of them be happy together, Claire starts showing up. They go to the movies, Claire wants to come too. They’re going to grab lunch, and Claire’s totally there. If they’re all at a party, you know that Claire is all over Daniel, and sitting with them, and doing everything she can to pop up often enough, to come between them often enough, to let Daniel see what she has to offer and to stop him from dating that ^%$@$ Kate!

That, my friends, is epenthesis. When one person pops up to try and prevent a combination or situation which is undesirable. Mind you, it can also occur in other social situations. If you find out your blind date is rather creepy, you might arrange to meet up with a friend later in the evening, so that your date won’t get any ideas and so your friend can stop any really bad interactions from occurring. Or it can even be so simple as choosing to tag along when her creepy ex-boyfriend offers to walk your intoxicated friend home. Each of these is a situation where a person pops up, seemingly out of nowhere, to prevent an awkward combination.

In Phonology, if sounds suddenly pop up, that needs to be your cue to be on the lookout for awkward situations. If there’s suddenly a glottal stop between those two vowels, maybe it’s a sign that the language doesn’t want those vowels to be together. Or if you suddenly get a schwa between two sounds which with otherwise assimilate, you can bet that the language isn’t interested in that assimilation occurring. Put differently, “OMG that /t/ TOTALLY can’t turn into a nasal next to that /n/ or it will ruin THE WHOLE MORPHEME! I’m totally gonna epenthesize a schwa to break them up. Besides, that /t/ shouldn’t be hooking up with that skank /n/ anyways, he’s MINE!”

#4: Deletion/Uninviting

Deletion in Phonology is, as you might imagine, when a sound gets deleted from a word or morpheme to prevent an awkward or undesirable combination or situation.

A really good example of this is in the word “fifth”. Nominally, there are two /f/ sounds in there (/fɪfθ/), but if we leave that second /f/ in there, we end up with an /fθ/ combination, which is just pretty awkward to say. So, usually, especially in faster speech, we’ll delete that second /f/ altogether (leaving us with [fɪθ]).

Deletion also happens a lot in borrowings. In Russian, the city we call “Moscow” is Москва (pronounced “Mosk-va”, /moskvɑ/). But we English speakers just don’t care for that /skv/ cluster, so we’ve chosen to delete the /v/ altogether (as well as taping on a diphthong), giving us our pronunciation, /mɑskaw/.


Lonely Panda
It’s OK, Panda, there are always other parties.

Deletion happens all the time in High School Drama. If you’re planning this totally awesome party and you invited Kirk and Sally and Jenny and Ron and all the cool kids, then suddenly Kirk and Sally break up, well, you’re in a quandary. If you invite both Kirk and Sally, they’ll be fighting and bickering and awkward and just being an undesirable combination, so you don’t want to do that. But if you uninvite them both, then everybody’s going to miss them, and you’ll seem really uncool. So you have to carefully weigh who would bring more to the party, and eventually choose which of them will be missed less, and uninvite them.

Similarly, if it looks like a sound is being deleted, your first question should be “why”? Why would it be awkward or undesirable for that sound to show up there? Obviously, /k/ and /v/ didn’t just break up with each other, but clearly we have a motivation to not let them get together. So, remember, no phonological change is random, and languages tend not to like deleting things without good cause, so that little deletion may represent avoidance of a really nasty situation for speakers. Because nobody wants awkward morphemes.

There’s always more drama

Of course, these are just the four main types of phonological rules and changes which occur, and there is no shortage of other interesting, more specific types of changes. But, taking a broad view, many of the phonological changes you’ll deal with throughout your linguistic career are going to result in assimilation, dissimilation, epenthesis or deletion. And as you’re puzzling over why that /i/ just disappeared or why that /x/ suddenly became an /ʃ/, I encourage you to consider the more human side of phonology, and to think back to high school and all the crazy social alternations which happened there.

Because NOBODY puts lax vowels at the end of MY syllables!!!!1!

Tagged with Conventional Linguistics, Language Humor, Phonetics and Phonology | 12 Comments

«Read newer posts — Read older posts »

All content © Linguistic Mystic 2006-2012 unless otherwise indicated.
Please contact me for permission to reprint or repost.

Site Information

  • Categories

  • Latest Non-linguistic Posts

  • Archives

    • Site features