<section data-background="img/puzzleworld.jpg"> <section style="background-color:rgba(255, 255, 255, 0.5)"> <h1>Linguistics 120: Morphology</h1> <h3>Dr. Will Styler - Spring 2019</h3> </section> --- ### Administrative Notes - My Friday Office Hours are cancelled - But I'll have an extra hour today, after class - I won't be able to respond to emails over the weekend - Amber will give a guest talk Friday - Clickers will be used! - Clicker scores will be finalized early next week along with SONA - If I've asked you to remind me to change anything, now's the time - Graduating seniors, come see me after class! --- ### Grades - There's a good chance that I'll be doing a bit of curving (by HW or overall) - Don't be too disheartened if your current grade is feeling low - Bad grades will still be possible, but I want good students to get good grades - The goal of difficult assignments is to challenge you, not destroy you --- ### Homework 4 is due Friday - **You'll need to work alone on this** --- ### Please fill out your CAPE evaluations - It makes a difference, for me and future students - Be honest, but also be fair --- ### Today's Plan - What more awaits you in Morphology - What we have learned - The secret to great linguistic analysis --- # The Morphology we missed --- ### Quarters are short - There's always a balancing act - "If I focus on this, I'll lose that" - "... but I want to talk about case!" - More time with data is good - ... but takes away time for theory - Morphology is so broad in interactions that you can't cover it all! --- ### There remain many topics for you in morphology! - (We had to save something for grad school, right?) - So, I'll give a sneak preview of some of the things we've missed --- ## Compositionality --- ### Compositionality - "Every complex expression can be exhaustively broken down into minimally meaningful pieces, with no units of form or meaning left unaccounted for” - Are there meanings that are 'unaccounted for'? - Are there words which have more forms than meanings? - Are there words which have more meanings than forms? - Are there words whose meaning cannot be predicted from the forms? --- ### Nahuatl to the rescuetl <img class="r-stretch wide" src="morphology/nahuatlnoncomp.jpg" alt="$txt"> --- ### It sure seems like there's non-compositional meaning! - 'Drink-water' becomes 'drown' - ... but only when incorporated! - 'Flower-make' becomes 'arrange flowers' - ... but only when incorporated! - 'deer-flee' becomes 'skittishly and quickly run away' --- ### "... but maybe those are just morphemes with different meanings!" - Well, OK, but how about... --- ## Phonaesthemes --- <section data-background-color=#708cef></section> ### "Would you like to buy a new Glimp? It..." a) Sharpens knives. b) Lights up your house. c) Tastes good. --- <section data-background-color=#708cef></section> ### "Damnit, I left my meb in the car!" a) I won't be able to sew as well. b) I'll be unable to fill my bird feeders c) What if there's a fire?! --- <section data-background-color=#708cef></section> ### "Have you seen my snofter?..." a) My nose is sore. b) I dropped a glass. c) It's cold in here. --- <section data-background-color=#708cef></section> ### "I got diabetes. It must be from drinking all that..." a) Blurge cola b) Mabler cola c) Wika cola --- <section data-background-color=#708cef></section> ### "You Mom is plafty." a) "Them's fighting words." b) "Maybe?" c) "Thanks!" --- <section data-background-color=#708cef></section> ### "Your Mom is slafty." a) "Them's fighting words." b) "Maybe?" c) "Thanks!" --- ## Phonaesthemes "Frequently recurring sound-meaning pairings that are not clearly contrastive morphemes" - First defined by Ben Bergen ([Bergen 2004](http://muse.jhu.edu/login?auth=0&type=summary&url=/journals/language/v080/80.2bergen.pdf)), who is at UCSD! --- ## Ben Bergen <img class="r-stretch tall" src="img/people_benbergen.jpg" alt="$txt"> --- ### Phonaesthemes in English - gl- ‘light, vision’ *glimmer, glisten, glitter, gleam, glow, glint*, etc. - sn- ‘nose, mouth’ *snore, snack, snout, snarl, snort, sniff, sneeze*, etc. - sl - 'pejorative', *slack, slouch, sludge, slime, slash, sloppy, slug, sluggard, slattern, slut, slang, sly, slither, slow, sloth*, etc. - b_l - 'expand abnormally', *blimp, bulge, bulk, balloon, blip* --- ### These are clearly not morphemes - In fact, they're part of other morphemes (e.g. "snack") - ... but they sure carry meaning! - These *emerged* as statistical accidents in the lexicon - "Lots of GL words mean light, so GL must mean light!" - They do end up seeming 'real' to speakers! - Neat! --- - Speaking of things that aren't morphemes... --- ## Lessthemes --- ### Nah, that's not really a thing. - ...but for half a second, you all were reanalyzing 'morpheme' pretty hard - <img class="r-stretch wide" src="humorimg/evillaugh.gif" alt="$txt"> --- Instead, let's check out... --- ## Multi-Word Expressions --- ### We talked about compounds - "Bleu Cheese Salad Dressing" - "Laptop" - "Deer-Flee" - "Inflectional Morphology" --- ### Compounds are continuous chunks! - "I'm going to Bleu Cheese the heck out of this Salad Dressing" is not a compound use - Although "blue cheese" and "salad dressing" may still be --- ### There are other 'words' which are discontinuous - "I threw my lunch up on the roof" - "I threw up my lunch on the roof" - "We're going to cut Mary out of the agreement" - "We're going to cut out Mary from the agreement" --- ### These verb-particle constructions feel like words in many ways - "Hit up" != "Hit" - "Throw out/up/in/back" - "Cut up/across/down/out/in/with" - They're separable, though - ... and comprised of multiple parts - These are super difficult for computers to handle --- ### Speaking of which... --- ## Computational Morphology --- ### Morphology complicates natural language processing - We need some way to teach Siri that 'walks' and 'walk' and 'walking' and 'walker' and 'walked' are the same - "I saw Cardi B's dog walk to the store" and "I saw Cardi B dog walk Tomi Lahren" - "Siri, remind me to hit Jamie up" - "OK, I'll remind you to hit Jamie." - "User has entered a word "swash". What's the plural of it?" --- ### Quantitative Research into Morphology is crucial - How can we solve morphological problems by language modeling - How can we use large amounts of data predict to morphological behavior? - What can huge corpora teach us about usage? - How does machine learning help us understand human learning? --- ### The interface between morphology, learning, and computer work will be Friday - Amber's got a great talk planned - Be there, or be super lame - Speaking of interfaces... --- ## Interfaces --- ### Morphology connects to everything! - Phonetics can be affected by lexical considerations - Morphophonology is 100% a thing - Morphology and Syntax interact at many levels - Semantics is the meaning half of form and meaning - Pragmatics is crucial for context and contrast --- ### There's much more waiting for you in Morphology! - That's why it's called 'More-phology!' - <img class="r-stretch wide" src="humorimg/badpundog.jpg" alt="$txt"> --- # Looking back at this quarter --- ### What did you learn about... - Morphological Analysis? - Morphological patterns? - What 'word' means? - How words might be created? - About the lexicon? --- ### You've learned a lot! - You can now solve morphological problems with elegance, generality, realism, and a minimum of complexity - You know how morphology works, around the world! - You know some ways of dealing with weirdly shaped 'words' and affixes! - You've got a good sense of ways too think about word creation and productivity! - You know the big questions about the lexicon, and a few different answers! --- ### We've also had some big discussions about the nature of language - Should we describe language through parts and rules, which *generate* the proper forms? - Should we describe language by looking at the *usage*, and the make generalizations as needed? - Should we take the middle path, and say that grammar may emerge, but the patterns are real, too? - Which of these is cognitively real? - ... does it even matter? --- ### ... and you learned some valuable life lessons - A good analysis isn't good if you can't understand it - Don't assume that anything must be (or mustn't be) marked - Don't assume that 'salt' isn't secretly a verb. - Don't assume that a new language will work like you're used to - Don't open a beer with a chainsaw, just because you can. - **The past tense of 'yeet' is 'yote', and the participle is 'yaughten'** --- ### I know what you're thinking... - "Wow, all that amazing knowledge, in just 10 weeks?" - **"What more could we possibly ask for?"** --- # The secret to great linguistic analysis --- ### You've come a long way in the program - LIGN 101, 110, 111 - Many of you have taken syntax and semantics - Some of you are about to graduate! --- ### In the future, you'll all need to do linguistic analysis - Perhaps as part of your teaching or research, if you decide to go the academic route - Maybe in industry or defense, if that's your path - Maybe for fun, on the weekends - ... and during the week --- ## You'll want to do *great* linguistic analysis - So, I'm going to finally tell you the secret --- ### Most linguistics students don't learn this until graduate school - Many not until a post-doc - Some, not until they're full professors - Some poor souls never figure it out! - **... but you'll know it today!** --- ### We'll end the class like we've lived the class, staring at data! - *To discover the secret to Great Linguistic Analysis, you'll need to figure out...* --- ### What is the Russian marker for 1SG.PRES for these verbs? <style type="text/css"> .tg {border-collapse:collapse;border-spacing:0;} .tg td{font-family:Arial, sans-serif;padding:10px 5px;border-style:solid;border-width:1px;overflow:hidden;word-break:normal;border-color:black;} .tg th{font-family:Arial, sans-serif;font-weight:normal;padding:10px 5px;border-style:solid;border-width:1px;overflow:hidden;word-break:normal;border-color:black;} .tg .tg-0pky{border-color:inherit;text-align:left;vertical-align:top} .tg .tg-0pk{font-weight:bold;border-color:inherit;text-align:left;vertical-align:top} </style> <table class="tg"> <tr> <th class="tg-0pk">Russian</th> <th class="tg-0pk">English</th> <th class="tg-0pk">Russian</th> <th class="tg-0pk">English</th> <th class="tg-0pk">Russian</th> <th class="tg-0pk">English</th> </tr> </b> <tr> <th class="tg-0pky">djelat</th> <th class="tg-0pky">'to do'</th> <th class="tg-0pky">pit</th> <th class="tg-0pky">'to drink'</th> <th class="tg-0pky">staret</th> <th class="tg-0pky">'to grow old'</th> </tr> <tr> <th class="tg-0pky">djelaju</th> <th class="tg-0pky">'I do'</th> <td class="tg-0pky">pju</td> <td class="tg-0pky">'I drink'</td> <th class="tg-0pky">stareju</th> <th class="tg-0pky">'I grow old'</th> </tr> <tr> <td class="tg-0pky">djelajet</td> <td class="tg-0pky">'(s)he does'</td> <td class="tg-0pky">pjet</td> <td class="tg-0pky">'(s)he drinks'</td> <td class="tg-0pky">starejet<br></td> <td class="tg-0pky">'(s)he grows old'</td> </tr> <!-- <tr> <td class="tg-0pky">tʃitaju</td> <td class="tg-0pky">'I read'</td> <td class="tg-0pky">risuju</td> <td class="tg-0pky">'I draw'</td> </tr> <tr> <td class="tg-0pky">tʃitajet</td> <td class="tg-0pky">'(s)he reads'</td> <td class="tg-0pky">risujet</td> <td class="tg-0pky">'(s)he draws'</td> </tr>--> </table> --- ## That's the secret, right there - That morpheme - As you go forward in your linguistic career, remember that... --- # The Secret to Great Linguistic Analysis was /ju/ all along! --- ### For Next Time - Amber will give an awesome talk about her work on French - You'll turn in HW4 Friday - **Graduating Seniors, come see me after class!** --- <huge>Thank you!</huge>