Browsed by
Category: Features

Remember, remember

Remember, remember

A lot of the work that linguists do involves taking a language as it is spoken at a particular time, finding generalizations about how it operates, and coming up with abstractions to make sense of them. In English, for example, we identify a category of ‘number’ (with possible values ‘singular’ and ‘plural’); and we do that because in many ways the relationship between cat and cats is the same as that between mouse and mice, man and men, and so on, meaning that it would be useful to treat all of these pairings as specific examples of a more general phenomenon. We can then make the further generalization that whatever this linguistic concept of ‘number’ really is, it is not only relevant to nouns but also to verbs, and to some other items too – because English speakers all know that this cat scratches whereas these cats scratch, and you can’t have any other combination like *these cat scratch.

A black cat wearing bat wings for Halloween
This bat scratches

Once you start looking, you discover layer upon layer of generalizations like these, and you need more and more abstractions in order to take care of them all. This all gives rise to a view of language as a kind of machine built out of abstract principles, all coexisting at the same time inside a speaker’s head. On that basis, we can ask questions like: are there any principles that all languages use? Does having pattern X always go along with having pattern Y? Are there any generalizations that you can easily come up with, but that turn out not to be found anywhere? What does all this tell us about human psychology?

But that is not the only approach to language we could take. While we can point to a general principle of English to explain what is wrong with these cat, there is no similar principle explaining why we refer to the meowing, purring, scratching creature as a cat in the first place. The word cat has nothing feline about it, and the fact that we use that sequence of sounds – rather than e.g. tac – is not based on some higher-level truth that applies for all English speakers right now: instead, the ‘explanation’ is rooted in the fact that this is the word we happened to inherit from earlier generations of speakers.

Portrait photo of General Burnside, featuring his famous sideburns
General Ambrose Burnside (1824-1881)

So studying the etymology of individual words serves as a good reminder that as well as an abstract, principled system residing in human minds, every language is also a contingent historical artefact, shaped by the peoples and cultures of the past.1 Nothing makes this more obvious than the continued existence of ordinary vocabulary items that commemorate individuals from centuries gone by – often without modern-day speakers even knowing it. In English, sandwiches are named after the Earl of Sandwich, wellingtons are named after the Duke of Wellington, and cardigans are named after the Earl of Cardigan; and the parallelism here says something about the locus of cultural influence in Georgian and Victorian Britain. More cryptically, sideburns owe their name to a General Burnside of the US Army, justly famed for his facial hair; algorithms celebrate the Persian mathematician al-Khwarizmi; and Duns Scotus, although a towering figure of medieval philosophy, now lives on in the word dunce popularized by his academic opponents.2

But which historical figure has had the greatest success of all in getting his name woven into the fabric of modern English? I reckon that, against all the odds, it could well be this Guy.

A close up of the face of Guy Fawkes, labelled Guido Fawkes, from a depiction of several conspirators together

While all English speakers are familiar with the word guy as an informal word corresponding to man, probably not that many know that it can be traced back to a historical figure from 400 years ago who, in a modern context, would be called a religious terrorist. Guy Fawkes was one of the conspirators in the ‘Gunpowder Plot’ of November 1605: with the aim of installing a Catholic monarchy, they planned to assassinate England’s Protestant king, James I, by blowing up Parliament with him inside. Fawkes was not one of the leaders of the conspiracy, but he was the one caught red-handed with the gunpowder; as a result, one cultural legacy of the plot’s failure is the celebration every 5th November (principally in the UK) of Guy Fawkes Night, which commonly involves letting off fireworks and setting a bonfire on which a crude effigy of Fawkes was traditionally burnt.

But how did the name of one specific Guy, for a while the most detested man in the English-speaking world, end up becoming a ubiquitous informal term applying to any man? The crucial factor is the effigy. It is unsurprising that this came to be called a Guy, ‘in honour’ of the man himself; but by the 19th century, the word was also being used to refer to actual men who dressed badly enough to earn the same label, in the way one might jokingly liken someone to a scarecrow (one British woman writing home from Madras in 1836 commented: ‘The gentlemen are all ‘rigged Tropical’,… grisly Guys some of them turn out!’). It is not a big step from there to using guy as a humorous and, eventually, just a colloquial word for men in general.3

Procession of a Guy (1864)

And of course the story does not stop there. While a guy is still almost always a man, for many speakers the plural guys can now refer to people in general, especially as a term of address. The idea that a word with such unambiguously masculine origins could ever be treated as gender-neutral has been something of a talking point in recent years, as in this article from The Atlantic about the rights and wrongs of greeting women with a friendly ‘hey guys’; but the fact that it is debated at all shows that it is happening. In fact, there is good reason to think that in some varieties of English, you-guys is being adopted as a plural form of the personal pronoun you: one piece of evidence is the existence of special possessive forms like your-guys’s, a distinctively plural version of your.

It is interesting to notice that the rise of non-standard you-guys, not unlike y’all and youse, goes some way towards ‘fixing’ an anomaly within modern English as a system: almost all nouns, and all other personal pronouns, have distinct singular and plural forms, whereas the standard language currently has the same form you doing double duty as both singular and plural. Any one of these plural versions of you might eventually win out, further strengthening the (already pretty reliable) generalization that English singulars and plurals are formally distinct. This just goes to show that the two ways of looking at language – as a synchronic system, and as a historical object – need to complement each other if we really want to understand what is going on. At the same time, it is fun to think of linguists of the distant future researching the poorly attested Ancient English language of the twenty-second century, and wondering where the mysterious personal pronoun yugaiz came from. Would anyone who didn’t know the facts dare to suggest that the second syllable of this gender-neutral plural pronoun came from the given name of a singular male criminal, executed many centuries before?

  1. For example, cat itself seems to be traceable back to an ancient language of North Africa, reflecting the fact that cats were household animals among the Egyptians for millennia before they became popular mousers in Europe. []
  2. Of course, it is no accident that all of these examples feature men. Relatively few women in history have had the opportunity to turn into items of English vocabulary; in fact, fictional female characters – largely from classical mythology – have had much greater success, giving us e.g. calypso, rhea and Europe. []
  3. A similar thing also happened to the word joker in the 19th century, though it didn’t get as far as guy: that suggests that sentences containing guy would once have had the same ring to them as Who’s this joker?; and then some joker turns up and says… []
Is twote the past of tweet?

Is twote the past of tweet?

Have you ever encountered the form twote as a past tense of the verb to tweet? It is something of a meme on Twitter, and a live example of analogy (and its mysteries). However surprising the form may sound if you have never encountered it, it has been the prescribed one for a long time:

Ten years later, the question popped up among a linguisty Twitter crowd, where a poll again elected twote as the correct form:

It is clear that this unusual form replacing tweeted is some sort of form, but why specifically twote? I saw here and there a reference to the verb to yeet, a slang verb very popular on the internet and meaning more or less “to throw”. Rather than a regular form yeeted, the past for to yeet is often taken to be yote. The choice of an irregular form is probably meant to produce a comedic effect.

This, precisely, is analogical production: creating a new form (twote) by extending a contrast seen in other words (yeet/yote). Analogy is a central topic in my research. I have been trying to answer questions such as: How do we decide what form to use ? How difficult is it to guess? How does this contribute to language change?

But first, have you answered the poll?

What is the past tense of “to tweet”?

To investigate further why we would say twote rather than tweeted, I took out my PhD software (Qumin). Based on 6064 examples of English verbs1, I asked Qumin to produce and rank possible past forms of tweet2. To do so, it read through examples to construct analogical rules (I call them patterns), then evaluated the probability of each rule among the words which sound like tweet.

Qumin found four options3: tweeted (/twiːtɪd/), by analogy with 32 similar words, such as greet/greeted; twet (/twɛt/), by analogy with words like meet/met; tweet (/twiːt/) by analogy with words like beat/beat, finally twote (/twəˑʊt/), by analogy with yeet. Figure 1 provides their ranking (in ascending order) according to Qumin, with the associated probabilities.

Twote 0.028 < tweet 0.056 < twet 0.056 < tweeted 0.86
Figure 1. Qumin’s ranking of the probability for potential past forms of to tweet

As we can see, Qumin finds twote to be the least likely solution. This is a reasonable position overall (indeed, tweeted is the regular form), so why would both the official Twitter account and many Twitter users (including several linguists) prefer twote to tweeted?

But Qumin has no idea what is cool, a factor which makes yeet/yote (already a slang word, used on the internet) a particularly appealing choice. Moreover, Qumin has no access to semantic similarity, which could also play a role. Verbs that have similar meanings can be preferred as support for the analogy. In the current case, both speak/spoke and write/wrote have similar pasts to twote, which might help make it sound acceptable. Some speakers seem to be aware of these factors, as seen in the tweet above.

What about usage?

Are most speakers aware of the variant twote and using it? Before concluding that the model is mistaken, we need to observe what speakers actually use. Indeed, only usage truly determines “what is the past of tweet”. For this, I turn to (automatically) sifting through Twitter data.

Speakers must choose between tweeted or twote: what a dilemna !

A few problems: first, the form “tweet” is also a noun, and identical to the present tense of the verb. Second, “twet” is attested (sometimes as “twett”), but mostly as a synonym for the noun “tweet” (often in a playful “lolcat” style), or as a present verbal form, with a few exceptions, usually of a meta nature (see tweets below). I couldn’t find a way to automatically distinguish these from past forms while also managing within the Twitter API limits. Thus, I left out both from the search entirely. This leaves only our two main contestants.

 

I extracted as many recent tweets containing tweeted or twote as Twitter would let me — around 300 000 tweets twotten between the 26th of August and the 3rd of September. 186777 tweets remained after refining the search4. Of these, less than 5000 contain twote:

There were more than 180000 occurences of tweeted and less than 5000 of twote in the past few days.
Counts of tweets containing either of two possible pasts for the verb “to tweet” in the past few days on twitter (mentions excluded).

As you can see, the tweeted bar completely dwarfs the other one. However amusing and fitting twote may be, and despite @Twitter’s prescription (but conforming with Qumin’s prediction), the regular past form is by far the most used, even on the platform itself, which lends itself to playful and impactful statements. This easily closes this particular English Past Tense Debate. If only it were always this simple!

  1. The English verb data I used includes only the present and past tenses, and is derived from the CELEX 2 dataset, as used in my PhD dissertation and manually supplemented by the forms for “yeet”. The CELEX2 dataset is commercial, and I can not distribute it. []
  2. The code I used for this blog post is available here, but not the dataset itself. Note that for scientific reasons I won’t discuss here, this software works on sounds, not orthography. []
  3. One last possibility has been ignored by this polite software, a form which follows the pattern of sit/sat. I see it used from time to time for its comic effect, but it does not seem at all frequent enough to be a real contestant (and I do not recommend searching this keyword on Twitter). []
  4. Since there has been a lot of discussion on the correct form, I exclude all clear cases of mentions. I count as mentions any occurrences wrapped in quotations, co-occurring with alternate forms, mentioning past tense, or with a hashtag. Moreover, with the forms in –ed, it is likely that the past participle would be identical, but for twote, the past participle could well be twotten. To reduce the bias due to the presence of more past participles in the usage of tweeted, I also exclude all contexts where the word is preceded by the auxiliary forms has, have, had, is, are, was, were, possibly separated by an adverb. []
History on the Ground

History on the Ground

Linguists spend most of the year stuck to the computer monitor: analyzing data, reading, or writing papers. But the time comes when we have to roll up our sleeves and find our sense of adventure. Personally, this is my favorite time of the year! Going on a linguistic field trip often involves living in a local community and immersing yourself in a completely different culture. You learn so much about the customs, traditions, beliefs… And, of course, you learn a lot about the language.

A sunny day, with a number of houses amongst trees on a hillside.
South-Eastern Serbia, one of our fieldwork destinations

Linguistic field trips are essential for researchers who work with poorly documented languages. We prepare questionnaires, design experiments, and go to the local community to collect data that we need for our research. Here, at SMG, we do fieldwork a lot. You can get a glimpse of this fascinating part of a linguist’s life in some of the previous posts. Check out, for example, this beautiful piece on Archi, this account of cultural and language diversity in the South Pacific, and the most recent post about intricate ways to express respect in  Vanuatu.

However, there seems to be a limitation. What should you do if you study the history of some phenomenon? If you are not only interested in how the system is now, but also in how it was before? We cannot jump into a time machine and reemerge in the 15th-century world to run our questionnaires there. So surely historical texts and comparative grammars are the only way to go, and fieldwork is not useful here… Or is it? Well, it turns out that a field trip can be very helpful for extrapolating historical data, but only if you are lucky with the location. Fortunately, I am!

A low-lying pile of stones on a roadside with a line of hills in the background
My lucky place

In the project “Declining case: Inflectional loss in progress”, my colleagues and I study dialects of Serbian and Bulgarian. These two languages have been posing linguists a headache for more than a century already. Although they are quite closely related and are spoken side by side, they have a lot of significant differences in the grammatical structure. To name some, (1) Bulgarian has articles (like English a and the), while Serbian does not, (2) Serbian uses infinitives, while Bulgarian does not, and (3) Serbian nouns have a fully-fledged morphological case system, while Bulgarian nouns do not inflect for case at all. People still argue about the exact reasons for this, but there is a general consensus that Bulgarian has undergone certain changes because it is located in the so-called Balkan linguistic area. A cool thing is that there is no sharp border between the innovative grammatical system of Bulgarian and the conservative system of Serbian. Rather, in the geographical zone on both sides of the Serbian-Bulgarian border, we see a variety of intermediate systems.

A map of eastern Serbia and Western Bulgaria, with an area of hatching straddling the border between the two countries

Let us see how it works on the example of the case inflection, which we study in our project. Cases are used in some languages to mark grammatical relations, such as subject or object. Serbian does it in this way, while Bulgarian uses prepositions instead. Take a look at this table, where the word ‘Cyprus’ appears in different contexts. See how in Serbian this word changes the ending depending on the context and in Bulgarian it keeps the same form? Just like in English!

Serbian Bulgarian Translation
vole Kipar xaresvat Kipâr ‘They like Cyprus’
stanovništvo Kipra naselenieto na Kipâr ‘The population of Cyprus’
pomažu Kipru pomagat na Kipâr ‘They help Cyprus’
upravljaju Kiprom upravljat Kipâr ‘They govern in Cyprus’

Overall, Serbian has six cases, while Bulgarian uses one general case form. So, what do we see in the transitional zone? Well, depending on where exactly we look, we find different systems. For example, in a more western part of the transitional area, we can meet a system where they use three cases, while in a more eastern part we can find a two-case system.

Serbian Transitional system 1 Transitional system 2 Bulgarian
Case 1 Case 1 Case 1 No case
Case 2 Case 2 Case 2
Case 3 Case 3
Case 4
Case 5
Case 6

What does it mean? It looks like standard Bulgarian at some point in its development lost its case marking on nouns completely, while the dialects in the transitional zone underwent this change to a smaller degree. The further west we move, the less this change affected the dialect. This situation created an unprecedented opportunity for us. We can go to different places in the transitional zone, compare their systems to each other, and use this comparison to create a historical model of the loss of case. We do not need a time machine, we have the different stages of this process living side by side today!

This summer, for example, I went to the municipality of Brus, which is located in Southern Serbia. There I witnessed the initial stages of case decline. People in Brus still use all six cases, but sometimes replace one with another, or insert a preposition in phrases where standard Serbian would not have it. While interviewing people, I learned about some fascinating traditions in this area. For example, one of the oldest customs at the wedding is to put an apple at the highest point in the backyard, and the groom has to hit the apple with a gun.  If he fails to do so, he is not going to get his bride!

An apple hanging by a thread from the bough of a tree.

Apparently, in earlier times, a wedding would last for several days and involve all sorts of rituals. Unfortunately, most of them are lost now. It would be so nice to see how a wedding was celebrated then! But for this, I am afraid, we do need a time machine.

A picture is worth a thousand words: Choosing images for psycholinguistic research

A picture is worth a thousand words: Choosing images for psycholinguistic research

Linguists need to come up with different ways of testing our theories of how particular languages in the world function. We generally rely on two main methods of data collection – linguistic elicitation and corpus collection. With linguistic elicitation a linguist asks a speaker of a language: ‘How do you say “Monty Python is really funny” in your language?’ But can we be sure that what the speaker said is naturalistic and not just a word for word translation?

Linguists need naturalistic data and can also record stories and conversations to build up a representative sample of a language (a corpus). This however takes a lot of time, effort and dedication on the part of both the linguist and the community of speakers of a language. It might even be that – after years of toil – the particular construction that a linguist wants to look at is under-represented with a dearth of examples in the corpus.

Thankfully, there is a happy medium! We can combine cognitive psychological techniques and targeted linguistic elicitation, to create scenarios where speakers produce naturalistic responses. Of course, this technique brings with it another set of problems entirely.

Psycholinguistic experiments need to be carefully designed and can’t be made up on the fly in response to something a speaker of a language says to you; this is drastically different to standard linguistic elicitation where one can continually come up with new sentences to check, while in the middle of working with a speaker of a language.

In our current research on optimal categorisation we aim to find out how different nouns are assigned to different classifiers in a group of six related Oceanic languages spoken in Vanuatu and New Caledonia. Each language has a different inventory size of classifying particles — from two to 23 — which are used in possessive constructions, and categorise the possession in terms of its use or functionality.

Here are a few examples from the Iaai language, spoken in New Caledonia, which has the largest inventory of classifiers in our sample of languages:

(1a)	a-n			wââ	(b)	hanii-ny		wââ
        FOOD.CLASSIFIER-his	fish 		CATCH.CLASSIFIER-his	fish
        ‘his fish (to eat)		        ‘his fish (which he caught)’
(2a)	a-n			koko	(b)	noo-n			koko
	FOOD.CLASSIFIER-his	yam		PLANT.CLASSIFIER-his	yam
	‘his yam (to eat)’			‘his yam plant’

We want to see whether or not a particular noun that refers to a particular entity can occur with different classifiers, like with the words for ‘fish’ and ‘yam’ in Iaai above. Also, how does a language with 23 classifiers function differently from a language with just two or three classifiers?

One way in which we can discover how the classifiers function in each language is to use a card sorting experiment. These experiments present speakers with entities in the form of pictures. Speakers are asked to sort them into different groups, first in a “free sort” where they can create groups on any basis they feel is relevant and important, and second, in a “structured sort” where they are asked to group entities according to which classifier they would use in a possessive construction. By doing this with lots of participants we can see individual speaker variation in language usage in one language and across languages and get a clear sense of if and how a language’s classifier system is influencing the way that speakers think about and process different entities.

Once we have decided on which nouns to test in a card sort experiment we have to find or make pictures that represent these images. Sadly I don’t have the artistic skills of Michelangelo and won’t be painting any masterpieces for the experiment! 

Choosing what type of image is trickier than it sounds as we are presented with an array of options.

First should we use simple line drawings of the images? The Noun Project has over 2 million small black and white line drawings. With such a choice of images we can find what we need. Here are some images of yams that I found on the site that we could use for our experiment.

These are great, and I know they are yams because I searched for images of yams on the website. But if I present these images to speakers I want them to tell me what they are. If the images aren’t instantly recognisable then participants will use different nouns to describe what they are seeing – is it a yam? A sweet potato? Manioc? Or some other entity? Actually, to tell you the truth, the third picture is actually a sweet potato! But it looks very similar to the first picture of a yam. Another problem is that these images can be quite abstract – and we can’t be sure that these symbolic representations of entities will be shared across different cultural and linguistic groups.

What about black and white pictures? – These are cheaper to print and easier to standardise. But we do not see the world in black and white and presenting entities as black and white pictures  may make it harder to identify  them, especially when the lightness of the background and the object of focus are similar. We need to be sure that the images we choose are easy to identify or else we can end up with problems of misidentification.

Another possibility is to remove the background of the image.  By doing this we can eliminate distractions and help the participant focus on the object in the image. However, the background is often key. Background information gives context that can influence how the speaker of a language perceives the entity in the image.

For instance, speakers may classify a fish that has been caught differently to a fish that is alive and swimming in the sea. The edible classifier is more likely with the former scenario, and a general classifier with the latter. But if we were to remove the background from both of these photos they would look strikingly similar! This leads us onto a very important question – what classifier would speakers of these languages use for a parrot if it was alive or dead?

So now we have decided to present images in colour and keep the background. But we must make sure that the background varies across different images. We don’t want participants to sort the entities into groups based on a colour or shape in the background or some other extraneous visual cue that may appear in several pictures!

For every psycholinguistic experiment that uses images there are multiple decisions that need to be made to figure out what type of image is required. The images we have chosen are specifically tailored to the nature of the languages we are studying to ensure that they are culturally relevant and thus identifiable.

For us, the pictures need to be realistic and represent the world around us — Sadly, we can’t take artistic licence with kangaroos and trampoline acts, as fun as that would be!

 

Drinkable houses, edible canoes and Trojan horses

Drinkable houses, edible canoes and Trojan horses

Michael Lotito, a French entertainer known as Monsieur Mangetout, became famous for his penchant for devouring objects that most would consider inedible. From bicycles and televisions to the most bizarre of all, a Cessna 150 light aircraft.

Though Monsieur Mangetout hailed from France, one might have thought that he was from the archipelago of Vanuatu. This small island nation is not only famous for being the most linguistically dense country in the world – with over 130 languages for a population of just a quarter of a million – but is also renowned for its intriguing possessive classifiers, which turn up in sentences when you talk about the things that you own, much like the possessive pronouns in English – my, your, hers etc. But in the Oceanic languages of Vanuatu these classifiers also tell us about how you will use the item that you own.

It took Michael Lotito two years to eat the Cessna 150!

The most common distinctions these classifiers make are between three types of possessions: ones that can be drunk, eaten and a residual classifier used when the more specific instances of eating and drinking aren’t needed. So, for example if you speak Paamese you can make a distinction between a coconut that you will drink, ani mak ‘my drinkable coconut’; one that you will eat the flesh of, ani ak ‘my edible coconut’; or one that you intend to sell, ani onak ‘my coconut for an unspecified use’.

But, more intriguingly, several languages of Central Vanuatu, spoken on the islands of Pentecost, Ambrym, Paama and Epi, use the food and drink classifiers for some rather strange items that one might not consider to be edible or drinkable — though of course Michael Lotito might beg to differ. The drink classifier in the language of North Ambrym covers a rather broad range of entities, including the obvious drinks such as water, tea, coffee and juice:

(1)	ma-n			we	/	ti	/	jus
	DRINK.CLASSIFIER-his	water		tea		juice
	‘his water/tea/juice’

But the classifier is also used with items that can’t be drunk, like the words bwelaye ‘cup’ or bwela ōl ‘coconut shell (used as a cup)’, but also im ‘house’, hul ‘mat’ and bulubul ‘hole’. And in the Sa language spoken on Pentecost island, the food classifier can also be used with the word bulbul ‘canoe’!

The languages of Central Vanuatu where houses can be drunk, except for Raga which likes to be different.

How do you drink a house? How do you eat a canoe? While Michael Lotito might well be able to eat canoes and drink houses, the people who speak these languages certainly do not! So what explanation can be given as to why and how these non-drinkable and non-edible entities are included within the semantic domain of drinks and food?

The words meaning cups and containers of liquids are included with the drink classifiers in some of these languages due to a process of semantic extension. This is when the coverage of the semantics of a classifier are extended to include entities that are frequently associated with the core meaning of that classifier. This type of semantic extension is known as metonymy, where the word for a container can be used instead of the word for what it contains – e.g. in English we can use the word ‘dish’ to refer not only to a plate, but also to its contents. It is not such a large cognitive step to associate drinks with cups, and that is why containers of liquids are now included in the drink classifier’s semantic domain. However, it is quite a large cognitive leap to think that houses are associated with drinks and canoes with food.

To explain how houses are now classified along with drinks and canoes with food we have to look into the history of the languages and how these languages have changed through time. This is of course quite a difficult endeavour considering that these languages have no literary traditions and are only now just starting to be written down. We cannot  consult old texts to see how the language used to be several hundred years ago as these don’t exist. Though limited records exist for a few languages going back to the mid 1800s, we mainly have to rely on comparing how related languages in the area differ and try to figure out how they got to be different.

Let’s start by looking at the language of Apma, spoken on Pentecost. The word for house, imwa, doesn’t occur with the drink classifier, but instead occurs in a different possessive construction where the owner is marked directly on the word for house, instead of on a classifier:

(2)	imwa=n		atsi
	house=his	person
	‘a person’s house’

This type of construction, called direct possession, normally occurs with possessions closely associated with the possessor, including body parts and kinship terms, but sometimes includes more intimate personal possessions as well. Now if we look at Apma’s neighbouring language, Ske, spoken to the south, the noun for house occurs with the drink classifier:

(3)	im	mwa=n			azó
	house	DRINK.CLASSIFIER=his	person
	‘a person’s house’

As you can see the word for house in Ske, which historically for the languages of Pentecost would have been imwa just like it is in Apma, has been split, where the first part im now means ‘house’, and speakers recognise the second part of imwa, namely mwa, as identical in form to the drink classifier. Speakers have now reanalysed the second part of the word for house as being the drink classifier, and now accept houses as being classified along with drinkable entities. A similar mechanism has occurred across several other languages of Central Vanuatu, and this is why houses are classified along with drinks.

Just what is a drinkable house anyway?

In most languages of Vanuatu this change didn’t occur and houses are either directly possessed or occur with the residual general classifier. But in a few other languages, the word for house developed into a distinct classifier that is different from the drink classifier. In the languages of Southern Vanuatu the word for house iimwa has now turned into a classifier for locations and places, and is distinct from the classifier for drinks — nɨmwɨ.

Now what about the edible canoes that I mentioned earlier? This strange occurrence happens in the language of Sa, also spoken on Pentecost island:

(4a)	a-k			anian		(b)	a-k			bulbul
	FOOD.CLASSIFIER-my	food               	FOOD.CLASSIFIER-my	canoe
	‘my food’					‘my canoe’

Historically, the word for canoe was waga in Proto Oceanic, and the word for bulbul was used for a specific type of canoe. Sometimes linguists get lucky and there can be historical documents that help show us the way. Miss Hardacre, a missionary living in northern Pentecost in the early part of the twentieth century, made a small dictionary of the Raga language. In this dictionary she recorded the generic-specific word pairing waga bulbul, ‘canoe type/raft’. Now in Sa, the original word for canoe, waga, underwent several sound changes until it ended up looking like the food classifier, where only the medial vowel /a/ was left! The new word for canoe was bulbul, whereas the old generic term, waga, merged into the food classifier. In other languages of the area, such as Raljago, spoken on Ambrym, a separate classifier for canoes and boats emerged, distinct from the food classifier. Thus, the food classifier is a, but the canoe classifier is ai.

Sometimes when a merger takes place, the noun that merges into a classifier acts as a Trojan horse. Looking back to the language of North Ambrym, where the drink classifier can occur with other nouns denoting houses, parts of houses, and mats. The word for house that originally merged into the drink classifier acts as a locus for semantic extension, opening a back door to other nouns that are semantically similar — those that are in the domain of houses — to enter into the drink classifier as well.

I think Michael Lotito would have felt at home speaking one of the Oceanic languages of Vanuatu. He might even have said of his Cessna 150ː

(5)	a-k			Cessna 150
	FOOD.CLASSIFIER-my	Cessna 150
	‘my edible Cessna 150’

Many thanks to Andrew Gray who runs the languages of Pentecost Island website and is my co-conspirator in turning this post into a journal article!

Optimal Categorisation: How do we categorise the world around us?

Optimal Categorisation: How do we categorise the world around us?

People love to categorise! We do this on a daily basis, consciously and subconsciously. When we are confronted with something new we try and figure out what it is by comparing it to something we already know. Say, for instance, I saw something flying through the air – I may think to myself that the object is a bird, or I may say it is a plane based on my previous experiences of birds and planes. Of course the object may turn out to be something completely new, perhaps even superman!

Is it a bird? Is it a plane? No it’s Superman!

Our love of classification runs deep in scientific enquiry. Botanists and zoologists classify plants and animals into different taxonomies. Even the humble linguist loves to classify – is this new word a noun or a verb? What about the new word zoodle that was recently added to the Merrriam Webster dctionary? Is it a thing? Or an action? Can I zoodle something or is it something I can pick up and touch? Well apparently zoodle is a noun which means ‘a long, thin strip of zucchini that resembles a string or narrow ribbon of pasta’. To be honest, I love eating zoodles, though until now I never knew what they were called!

The way people classify entities around them has become encoded in the different languages we speak in many different ways. The most obvious example that springs to mind is when we learn a new language, like French or German, we are confronted with a grammatical gender system. French has two genders – Masculine and Feminine. But German has three – Masculine, Feminine and Neuter. Other languages can have many more gender distinctions. Fula, a language spoken in west and central Africa, has twenty different gender categories!

So what exactly are grammatical gender systems and how are they realised in different languages? Gender systems categorise nouns into different groups and tend to appear not on the noun itself, but on other elements in the phrase. In German, nouns are split into three different gender categories – masculine, feminine and neuter. The gender of a noun is shown by using different articles (the word ‘the’ or ‘a’) and sometimes by changing the ending of an adjective, but never on the noun itself. Thus the word for ‘the’ in German is either der, die or das depending on whether the noun in the phrase is masculine, feminine or neuter.

(1)        der       Mann
              the       man

(2)        die        Frau
              the       woman

(3)        das       Haus
              the       house

This is called ‘agreement’ as the adjectives and articles must agree with the gender of the noun. In a language with gender, each noun typically can only occur in one gender category.

Not every language has a grammatical gender system, but they are highly pervasive, with around 40% of all languages having such a system. English is quite a poor example when it comes to gender. There is no real gender agreement in English, with the exception of pronouns. We have to say: Bill walked into the grocers. He bought some apples. Where the pronoun he must agree with the gender of the noun that was previously mentioned. English uses he, she and it as the only markers of gender agreement.

Languages behave differently in how they allocate nouns to the different genders, which can be very baffling for language learners! Why in French is chair feminine, la chaise, but in German it is masculine, der Stuhl? How a language allocates nouns to its gender categories can seem somewhat arbitrary – with the exception of the words for women and men, which fall into the feminine and masculine genders being the only semantically obvious choices.

But wait! If you thought the English gender system was dull, think again! A couple of months ago my piano was being restored and when it was being moved back into the lounge the piano movers kept saying: “pull her a little bit more” and “turn her this way”. The movers used the female pronouns to describe the piano. In English, countries, pianos, ships and sometimes even cars use the feminine pronouns.

Grammatical gender isn’t the only way languages classify nouns. Some languages use words called classifiers to categorise nouns. Classifiers are similar to English measure terms, which categorise the noun in terms of its quantity, such as ‘sheet of paper’ vs. ‘pack of paper’ or ‘slice of bread vs. ‘loaf of bread’. Classifiers are found in languages all over the world and are able to categorise nouns depending on the shape, size, quantity or use of the referent, e.g. ‘animal kangaroo’ (alive) vs. ‘meat kangaroo’ (not alive). Classifier systems are very different to gender systems as nouns in a language with classifiers can appear with different classifiers depending on what property of the noun you wish to highlight. There are many different types of classifier systems, but to keep things short I am just going talk about possessive classifiers, which are mainly found in the Oceanic languages, spoken in the South Pacific.

When an item is in your possession we use possessive pronouns in English to say who the item belongs to. For instance if I say ‘my coconut’ – the possessive pronoun is my. In many Oceanic languages a noun can occur with different forms for the word my depending on how the owner intends to use it. For instance the Paamese language, spoken in Vanuatu, has four possessive classifiers and I could use the ‘drinkable’ if I was talking about my coconut that I was going to drink. I would use the ‘edible’ classifier if I was going to eat my coconut. I would use the classifier for ‘land’ if I was talking about the coconut growing in my garden. Finally, I could use the ‘manipulative’ classifier if I was going to use my coconut for some other purpose – perhaps to sit on!

(4)        ani                   mak
              coconut           my.drinkable
              ‘my coconut (that I will drink)’

(5)        ani                   ak
              coconut           my.edible
              ‘my coconut (that I will eat)’

Why do languages have different ways of categorising nouns? How do these systems develop and change over time? Are gender systems easier to learn than classifier systems? Are gender and classifiers completely different systems? Or is there more similarity to them than meets the eye? These are some of the big questions in linguistics and psychology. We are excited to start a new research project at the Surrey Morphology Group, called optimal categorisation: the origin and nature of gender from a psycholinguistic perspective, that seeks to answer these fundamental questions. Over the next three years we will talk more about these fascinating categorisation systems, explain our experimental research methods, introduce the languages and speakers under investigation, and share our findings via this blog. Just look out for the ‘Optimal Categorisation’ headings!

No we [kæn]

No we [kæn]

If something bad happened to someone you hold in contempt, would you give a fig, a shit or a flying f**k? While figs might be a luxury food item in Britain, their historical status as something that is valueless or contemptible puts them on the same level as crap, iotas and rats’ asses for the purposes of caring.

In English, we have a wide range of tools for expressing apathy. But we don’t always agree on how to express it, and even use seemingly opposite affirmative and negative sentences to express very similar concepts.  Consider the confusing distinction between ‘I couldn’t care less’ vs. ‘I could care less’ which are used in identical contexts by British and American speakers of English to mean pretty much the same thing. This mind-boggling pattern makes sense when we realise that those cold-hearted people who couldn’t care less have a care-factor of zero, while the others don’t care much, but could do so even less, if necessary.

Putting aside such oddities, negation is normally crucial to interpreting a sentence – words like ‘not’ determine whether the rest of the sentence is affirmative or negative (i.e. whether you’re claiming it is true or false). Accordingly, languages tend to mark negation clearly, sometimes in more than once place within a sentence. One of the world’s most robust languages in this respect is Bierebo, an Austronesian language spoken in Vanuatu, where no less than three words for expressing negation are required at once (Budd 2010: 518):

Mara   a-sa-yal              re         manu  dupwa  pwel.
NEGl   3PL.S-eat-find   NEG2  bird     ANA      NEG3
‘They didn’t get to eat the bird.’

While marking negation three times might seem a little inefficient, this pales in comparison to the problems that arise when you don’t clearly indicate it all. We only have to turn to English to see this at work, where the distinction between Received Pronunciation can [kæn] and can’t [kɑ:nt] is frequently imperceptible in American varieties where final /t/ is not released, resulting in [kæn] or [kən] in both affirmative and negative contexts.

You might think that once a word or affix or sound that indicates negation has been removed from a word, there isn’t anywhere else to go. But some Dravidian languages spoken in India really push the boat out in this respect. Instead of adding some sort of negative word or affix to an affirmative sentence to signal negation, the tense affix (past –tt or future -pp) is taken away, as shown by the contrast between literary Tamil affirmatives and negatives.

pati-tt-ēn                    pati-pp-ēn                  patiy-ēn
‘I learned’                  ‘I will learn.’               ‘I do/did/will not learn.’

This is highly unusual from a linguistic point of view, and it’s tempting to think that languages avoid this type of negation because it is difficult to learn or doesn’t make sense design-wise. But historical records show similar patterns have been attested across Dravidian languages for centuries. This demonstrates that inflection patterns of this kind can be highly sustainable when they come about – so we might be stuck with the can/can’t collapse for a while to come.

Today’s vocabulary, tomorrow’s grammar

Today’s vocabulary, tomorrow’s grammar

If an alien scientist were designing a communication system from scratch, they would probably decide on a single way of conveying grammatical information like whether an event happened in the past, present or future. But this is not the case in human languages, which is a major clue that they are the product of evolution, rather than design. Consider the way tense is expressed in English. To indicate that something happened in the past, we alter the form of the verb (it is cold today, but it was cold yesterday), but to express that something will happen in the future we add the word will. The same type of variation can also be seen across languages: French changes the form of the verb to express future tense (il fera froid demain, ‘it will be cold tomorrow’, vs il fait froid aujourd’hui, ‘it is cold today’).

The future construction using will is a relatively recent development. In the earliest English, there was no grammatical means of expressing future time: present and future sentences had identical verb forms, and any ambiguity was resolved by context. This is also how many modern languages operate. In Finnish huomenna on kylmää ‘it will be cold tomorrow’, the only clue that the sentence refers to a future state of affairs is the word huomenna ‘tomorrow’.

How, then, do languages acquire new grammatical categories like tense? Occasionally they get them from another language. Tok Pisin, a creole language spoken in Papua New Guinea, uses the word bin (from English been) to express past tense, and bai (from English by and by) to express future. More often, though, grammatical words evolve gradually out of native material. The Old English predecessor of will was the verb wyllan, ‘wish, want’, which could be followed by a noun as direct object (in sentences like I want money) as well as another verb (I want to sleep). While the original sense of the verb can still be seen in its German cousin (Ich will schwimmen means ‘I want to swim’, not ‘I will swim’), English will has lost it in all but a few set expressions like say what you will. From there it developed a somewhat altered sense of expressing that the subject intends to perform the action of the verb, or at least, that they do not object to doing so (giving us the modern sense of the adjective ‘willing’). And from there, it became a mere marker of future time: you can now say “I don’t want to do it, but I will anyway” without any contradiction.

This drift from lexical to grammatical meaning is known as grammaticalisation. As the meaning of a word gets reduced in this way, its form often gets reduced too. Words undergoing grammaticalisation tend to gradually get shorter and fuse with adjacent words, just as I will can be reduced to I‘ll. A close parallel exists in in the Greek verb thélō, which still survives in its original sense ‘want’, but has also developed into a reduced form, tha, which precedes the verb as a marker of future tense. Another future construction in English, going to, can be reduced to gonna only when it’s used as a future marker (you can say I’m gonna go to France, but not *I’m gonna France). This phonetic reduction and fusion can eventually lead to the kind of grammatical marking within words that we saw with French fera, which has arisen through the gradual fusion of earlier  ferre habet ‘it has to bear’.

Words meaning ‘want’ or ‘wish’ are a common source of future tense markers cross-linguistically. This is no coincidence: if someone wants to perform an action, you can often be reasonably confident that the action will actually take place. For speakers of a language lacking an established convention for expressing future tense, using a word for ‘want’ is a clever way of exploiting this inference. Over the course of many repetitions, the construction eventually gets reinterpreted as a grammatical marker by children learning the language. For similar reasons, another common source of future tense markers is words expressing obligation on the part of the subject. We can see this in Basque, where behar ‘need’ has developed an additional use as a marker of the immediate future:

ikusi    behar   dut

see       need     aux

‘I need to see’/ ‘I am about to see’

This is also the origin of the English future with shall. This started life as Old English sceal, ‘owe (e.g. money)’. From there it developed a more general sense of obligation, best translated by should (itself originally the past tense of shall) or must, as in thou shalt not kill. Eventually, like will, it came to be used as a neutral way of indicating future time.

But how do we know whether to use will or shall, if both indicate future tense? According to a curious rule of prescriptive grammar, you should use shall in the first person (with ‘I’ or ‘we’), and will otherwise, unless you are being particularly emphatic, in which case the rule is reversed (which is why the fairy godmother tells Cindarella ‘you shall go to the ball!’). The dangers of deviating from this rule are illustrated by an old story in which a Frenchman, ignorant of the distinction between will and shall, proclaimed “I will drown; nobody shall save me!”. His English companions, misunderstanding his cry as a declaration of suicidal intent, offered no aid.

This rule was originally codified by Bishop John Wallis in 1653, and repeated with increasing consensus by grammarians throughout the 18th and early 19th centuries. However, it doesn’t appear to reflect the way the words were actually used at any point in time. For a long time shall and will competed on fairly equal terms – shall substantially outnumbers will in Shakespeare, for example – but now shall has given way almost entirely to will, especially in American English, with the exception of deliberative questions like shall we dance? You can see below how will has gradually displaced shall over the last few centuries, mitigated only slightly by the effect of the prescriptive rule, which is perhaps responsible for the slight resurgence of shall in the 1st person from approximately 1830-1920:

Until the eventual victory of will in the late 18th century, these charts (from this study) actually show the reverse of what Wallis’s rule would predict: will is preferred in the 1st person and shall in the 2nd , while the two are more or less equally popular in the 3rd person. Perhaps this can be explained by the different origins of the two futures. At the time when will still retained an echo of its earlier meaning ‘want’, we might expect it to be more frequent with ‘I’, because the speaker is in the best position to know what he or she wants to do. Likewise, when shall still carried a shade of its original meaning ‘ought’, we might expect it to be most frequent with ‘you’, because a word expressing obligation is particularly useful for trying to influence the action of the person you are speaking to. Wallis’ rule may have been an attempt to be extra-polite: someone who is constantly giving orders and asserting their own will comes across as a bit strident at best. Hence the advice to use shall (which never had any connotations of ‘want’) in the first person, and will (without any implication of ‘ought’) in the second, to avoid any risk of being mistaken for such a character, unless you actually want to imply volition or obligation.

Words apart: when one word becomes two

Words apart: when one word becomes two

As any person working with language knows, the list of words from which we build our sentences is not a fixed one but rather is in a state of constant flux. Words (or lexemes in linguists’ terminology) are constantly being borrowed (such as ‘sauté’ from French), coined (such as ‘brexit’ from a blend of ‘Britain’ and ‘exit’) or lost (such as ‘asunder’, a synonym for ‘apart’). These happen all the time. However, two more logical processes exist that can alter the total number of entries in the dictionary of our language. Occasionally, lexemes may also merge, if two or more become one; or split, if one becomes two. These more exotic cases constitute a window into the fascinating workings of the grammar. In this blog I will present the story of one of these splitting events. It involves the Spanish verb saber, from Latin sapiō.

The verb’s original meaning must have been ‘taste’ in the sense of ‘having a certain flavour’, as in the sentence “Marmite tastes awful”. At some point it also began to be used figuratively to mean ‘come to know something’, not only by means of the sense of taste but also for knowledge arrived at by means of other senses. It is interesting that in the Germanic languages it seems that it was sight rather that taste that was traditionally used in the same way. Consider, for instance, the common use, in English, of the verb ‘see’ in contexts like “I see what you mean”, where it is interchangeable with ‘know’. Whether the source verb can be explained by the differences between traditional Mediterranean and Anglo-Saxon cuisines I’d rather not suggest for fear of deportation.

In any case, what must have been once a figurative use of the verb ‘taste’ became at some point the default way of expressing ‘know’. These are the two main senses of saber in contemporary Spanish and of its equivalents in most other Romance languages. The question I ask here is: do speakers of Spanish today categorize this as one word with two meanings? Or do they feel they are two different words that just happen to sound the same? There may be a way to tell.

In Spanish, unlike in English, a verb can take dozens of different forms. The shape of a verb changes depending on who is doing the action of the verb, whether the action is a fact or a wish etc. Thus, for example, speakers of Spanish say yo sé ‘I know’ but t sabes ‘you know’. They also use one form (so-called ‘indicative’) in sentences like yo veo que t sabes inglés ‘I see that you know English’ but a different form (so-called ‘subjunctive’) in yo espero que t sepas inglés ‘I hope that you know English’. The Real Academia Española, the prescriptive authority in the Spanish language, has ruled that, because saber is a single verb, it should have the same forms (sé, sabes etc.) regardless of its particular sense. Speakers, however, have trouble to abide by this rule, which is probably the reason why the need for a rule was felt in the first place. My native speaker intuition, and that of other speakers of Spanish, is that the verb may have a different form depending on its sense:

Forms of Spanish saber (forms starting with sab– in light gray, forms starting with sep– in dark gray)

The most obvious explanation for why this change could happen is that, when the two main senses of saber drifted sufficiently away from each other, speakers ceased to make the generalization that they were part of the same lexeme. When this happened, the necessity to have the same forms for the two meanings of saber dissappeared. But, why sepo?

Because cannibalism is on the wane (also in Spain) we hardly ever speak about how people taste. As a result, the first and second person forms of saber (e.g. irregular ) are only ever encountered by speakers under their meaning ‘know’. Because of this, they do not count as evidence for language users’ deduction of the full array of forms of saber. This meant that the first and second person forms of saber₂ ‘taste’, when needed (imagine someone saying sepo salado ‘I taste salty’ after coming out of the sea), had to be formed on the fly on evidence exclusive to its sense ‘taste’ (i.e. third persons and impersonal forms):

Because of the evidence available to speakers, at first sight it might seem strange that this ‘fill-in-the-gaps’ exercise did not result in the apparently more regular 1SG indicative form sabo. This would have resulted in a straightforward indicative vs subjunctive distinction in the stem. The chosen form, however, makes more sense when one observes the patterns of alternation present in other Spanish verbs:

Verbs that have a difference in the stem in the third person forms between indicative and subjunctive (cab- vs quep- or ca- vs caig-) overwhelmingly use the form of the subjunctive also in the formation of the first person singular indicative. This is a quirk of many Spanish verbs. It appears that, by sheer force of numbers, the pattern is spotted by native speakers and occasionally extended to other verbs which, like saber look like could well belong in this class.

In this way, the tiny change from to sepo allows us linguists to see that patterns like those of caber and caer are part of the grammatical knowledge of speakers and are not simply learnt by heart for each verb. In addition, it gives us crucial evidence to conclude that, today, there are in Spanish not one but two different verbs whose infinitive form is saber. Much like the T-Rex in Jurassic Park, we linguists can sometimes only see some things when they ‘move’.

A plurality of plurals

A plurality of plurals

Of all the world’s languages, English is the most widely learnt by adults. Although Mandarin Chinese has the highest number of speakers overall, owing to the huge size of China’s population, second-language speakers of English outnumber those of Mandarin more than three times.

Considering that the majority of English speakers learn the language in adulthood, when our brains have lost much of their early plasticity, it’s just as well that some aspects of English grammar are pretty simple compared to other languages. Take for example the way we express the plural. With only a small number of exceptions, we make plurals by adding a suffix –s to the singular. The pronunciation differs depending on the last sound of the word it attaches to – compare the ‘z’ sound at the end of dogs to the ‘s’ sound at the end of cats, and the ‘iz’ at the end of horses – but it varies in a consistently predictable way, which makes it easy to guess the plural of an English noun, even if you’ve never heard it before.

That’s not the case in every language. Learners of Greek, for example, have to remember about seven common ways of making plurals. Sometimes knowing the final sounds of a noun and its gender make it possible to predict the plural, but  other times learners simply have to memorise what kind of plural a noun has: for example pateras ‘father’ and loukoumas ‘doughnut’ both have masculine gender and singulars ending in –as, but in Standard Greek their plurals are pateres and loukoumathes respectively.

This is similar to how English used to work. Old English had three very common plural suffixes, -as, -an and –a, as well as a number of less common types of plural (some of these survive marginally in a few high-frequency words, including vowel alternations like tooth~teeth and zero-plurals like deer). The modern –s plural descends from the suffix –as, which originally was used only for a certain group of masculine nouns like stān, ‘stone’ (English lost gender in nouns, too, but that’s a subject for another blog post).

How did the -s plural overtake these competitors to become so overwhelmingly predominant in English? Partly it was because of changes to the sounds of Old English as it evolved into Middle English. Unstressed vowels in the last syllables of words, which included most of the suffixes which expressed the gender, number and case of nouns, coalesced into a single indistinct vowel known as ‘schwa’ (written <ə>, and pronounced like the ‘uh’ sound at the beginning of annoying). Moreover, final –m came to be pronounced identically to –n. This caused confusion between singulars and plurals: for example, Old English guman ‘to a man’ and gumum ‘to men’ both came to be pronounced as gumən in Middle English. It also caused confusion between two of the most common noun classes, the Old English an-plurals and the a-plurals. As a result they merged into a single class, with -e in the singular and -en in the plural.

This left Middle English with two main types of plural, one with –en and one with –(e)s. Although a couple of the former type remain to this day (oxen and children), the suffix –es was gradually generalised until it applied to almost all nouns, starting in the North of England and gradually moving South.

A similar kind of mass generalisation of a single strategy for expressing a grammatical distinction is often seen in the final stages of language death, as a community of speakers transition from a minority to a majority language as their mother tongue. Nancy Dorian has spent almost 50 years documenting the dying East Sutherland dialect of Scots Gaelic as it is supplanted by English in three remote fishing villages in the Scottish highlands. In one study the Gaelic speakers were divided into fluent speakers and ‘semi-speakers’, who used English as their first language and Gaelic as a second language. Dorian found that the semi-speakers tended to overgeneralise the plural suffix –an, applying it to words for which fluent speakers would have used one of another ten inherited strategies for expressing plural number, such as changing the final consonant of the word (e.g. phũ:nth ‘pound’, phũnčh ‘pounds’), or altering its vowel (e.g. makh ‘son’, mikh ‘sons’).

But why should the last throes of a dying language bear any resemblance to the evolution of a thriving language like English? A possible link lies in second language acquisition by adults. At the same time as these changes were taking place, English was undergoing intense contact with Scandinavian settlers who spoke Old Norse. During the same period English shows many signs of Old Norse influence. In addition to many very common words like take and skirt (which originally had a meaning identical to its native English cognate shirt), English borrowed several grammatical features of Scandinavian languages, such as the suffix –s seen in third person singular present verbs like ‘she blogs’ (the inherited suffix ended in –th, as in ‘she bloggeth’), and the pronouns they, their and them, which replaced earlier hīe, heora and heom. Like the extension of the plural in –s, these innovations appeared earliest in Northern dialects of English, where settlements of Old Norse speakers were concentrated, and gradually percolated South during the 11th to 15th centuries.

It’s possible that English grammar was simplified in some respects as a consequence of what the linguist Peter Trudgill has memorably called “the lousy language-learning abilities of the human adult”. Research on second-language acquisition confirms what many of us might suspect from everyday experience, that adult learners struggle with inflection (the expression of grammatical categories like ‘plural’ within words) and prefer overgeneralising a few rules rather than learning many different ways of doing the same thing. In this respect, Old Norse speakers in Medieval England would have found themselves in a similar situation to semi-speakers of East Sutherland Gaelic – when confronted with a number of different ways of expressing plural number, it is hard to remember for each noun which kind of plural it has, but simple to apply a single rule for all nouns. After all, much of the complexity of languages is unnecessary for communication: we can still understand children when they make mistakes like foots or bringed.