The Other Pronouns

There’s been a lot of focus, including in my last post, on the question of pronouns recently, leading to peculiar responses such as where people say they “don’t have pronouns” or “don’t want to use pronouns”. This is weirdly ignorant but possibly reflects too strong a focus on a particular aspect of pronouns. Because of the way English works, most of this focus is on the third person and singular number versus the plural number, because English pronouns only explicitly express what we think of as gender in the third person. It could be said that there is some gendering of other pronouns, for instance I wouldn’t be surprised if married women are more likely to say “we” when referring to themselves than married men are, but the fact remains that we don’t perceive this variation much.

Today, though, I want to focus on the other pronouns, both personal and otherwise, because they tend to be lost in the heat of battle but are nonetheless interesting. These other pronouns, though not gendered in English, often are in other languages, but their gender is not the main thing I want to mention, because, to quote myself, “it’s the least interesting thing about them”. Well, usually. I will actually start with that though, and with the personal pronouns.

The basic system many languages have of personal pronouns is that they have singular and plural each of the first, second and third person: I, we, thou, you, it and they. English is unusual in this respect in that it lacks “thou”, more or less. Many languages have distinct polite and informal versions of “you” and distinct singular and plural versions of “you”, which can overlap such that plural “you” is often also polite “you”.

Urgh. I said I wasn’t going to discuss gender but I will because it’s worth getting it out of the way. In English we’re used to the third person singular pronouns being kind of gendered but actually not really. I’ll demonstrate. If for some reason we wanted to use “she”, “it” and “he” with an adjective, it wouldn’t vary according to that pronoun. Attributive adjectives are ungrammatical in English, but predicatives are very common, although in many Western European languages those don’t vary for gender. So, we say “she is tall” and “he is short”, but in French we’d add an “-e” for the feminine adjective. This happens, so far as I can tell, only with hair colour in English: a woman is blonde or brunette but a man is blond or brunet, and to be honest if I ever see the word “brunet” written down I shall be very surprised indeed. It doesn’t even extend to other hair colours such as “white”, “red”, “auburn” or whatever. It’s also annoying because it defines women by hair colour. This is basically the only time we can be even remotely said to use grammatical gender. Oh, actually I’m wrong: it crops up in the fossilised phrase “lady chapel” because that actually means the chapel belonging to The Lady, and is not “lady’s chapel”, so that’s a gender distinction. That’s it though, and in fact most people would probably perceive “brunette” and “brunet” as different words rather than the same word with a different ending.

One distinction we lack in English is gender in the second and first pronouns. This occurs in Spanish but in that it almost feels like an afterthought, and doesn’t occur in all cases. In other languages they generally seem to be fully-fledged pronouns as simple and short as each other, as opposed to having adjectives appended to them as in Spanish. As I’ve said, I won’t dwell on this.

In Old English, and even into early Middle English, there were dual personal pronouns, to refer to two people, as “wit”/”unc” and “git” (pronounced “yit”)/”inc”, the possessives being “uncer” and “incer”. Dual pronouns reappear in Tok Pisin, the Pidgin English of Papua, in a different form, and in Bislama, spoken in Vanu Atu, there’s also a trial form, but there’s a further complication with these so I won’t mention them yet. Gothic dual personal pronouns were “wit” and “jut”. In modern Icelandic, although the form of the dual pronouns survive, their sense doesn’t. They’re plural, and I think the original plural forms are now the polite forms of the pronouns but I might have misremembered. I personally think dual personal pronouns are very useful and I still use them in my diary as it often seems weird to use a plural for when there are only two of something. Dual third person pronouns have not been found in any recorded Germanic language, or in Latin. In Gothic, the verbs also have a dual conjugation for the first and second person, but this has never been found in English.

Something I meant to say last time but didn’t is that the operation of plural nouns sometimes looks quite like gender in English. “Scissors”, “glasses” and “trousers” for example refer to singular objects but use the plural. It seems slightly odd that “bra” is singular, since it could be “brum”, with “bra” as a plural if the etymology is ignored. These uses of the plural for singular objects involves “they” and only seems to happen when there is a duality to the item in question, so it’s like a trace of the dual although in fact it isn’t because there are genuine traces of the dual in English and even in modern pronouns, which again I’ll come to.

English has separate personal and demonstrative third person pronouns. Some other languages combine the two. The demonstratives are “this”, “that”, “these” and “those” and correspond in English also to the older words for “the” in the non-instrumental sense. Latin, Spanish and many other languages have three demonstratives, corresponding to the three persons, as in “hoc” (this by me), “iste” (that by you) and “id” (that over there). English gets by with just two, “this” and “that”. Remarkably, I think anyway although I’ve never come across anyone else mentioning it, Gothic has only one demonstrative in spite of generally making the finest distinctions of all Germanic languages: “þata”, which means both “this” and “that”. It’s the only language I’m aware of which only has one. Spanish does the same as Latin, with “eso”, “esto” and “aquel”, but as with some other Romance languages the referents have changed somewhat. There’s a language in Papua whose demonstrative pronouns refer to things like “towards the mountain”, “towards the lake” and so forth and there are very many of them. I thought this was Alamblak but apparently that just does the same as Spanish and Latin although its counting is peculiar, being based on 1, 2, 5 and 20 multiplied and added in various ways.

The real ‘Flowers For Algernon‘ personal pronouns for me are the inclusive and exclusive first person plurals. I found out about this distinction when I was about eleven and ever since it’s felt like a niggling but major problem for the English language. Remarkably, the vast majority of Indo-European languages get along without the distinction. It’s very simple, although I should warn you, you can’t unsee it once you know: many languages distinguish between “we but not you” and “you and I”. Austronesian languages such as Indonesian make this distinction, as do Dravidian languages. In the former case they also have an extra, dual number, meaning that there are five first person pronouns as opposed to the English two. I honestly can’t understand why we haven’t got these. It has been noted, though, that Indo-European words for “we” fall into two categories: ones of the “we” form and ones of the “nous” form. This suggests that clusivity did once exist in Proto-Indo-European but it didn’t even survive until the earliest form of the separate branches. Under the influence of Dravidian, some Indian languages do have this distinction although it isn’t related to these forms. I would say Papuan languages have this, but the thing about them is that there are hundreds and they vary a lot, so it’s possible to find many features in them which are present sporadically throughout the world.

This next bit is a bit mind-boggling in a peculiar way. It is technically possible for “you” to be inclusive or exclusive! This is interestingly difficult to think about. When one talks to someone, one says “thou” or “you”, but the “you”, being dual or plural, could refer to both or all the people one is addressing or it could refer to the people present and also to people not present. For instance, one could talk to a person and their partner, or one could talk to a person as part of the couple when the other person isn’t present. This apparently never happens though, even though it’s possible, and it’s thought by some linguists that this category of personal pronoun is impossible for the human mind to conceive of sufficiently clearly to exist. This raises further questions as to the nature of language. One linguist claims that this distinction is present in a critically-endangered language spoken in Vanu Atu called Southeast Ambryn, spoken by about three thousand people. I don’t think it’s that one can’t conceive of it. To me it seems simpler than split ergativity to get my head round. It’s more that someone who knows such a language would have to keep doing it, and as such it might inform issues around pronouns and gender identity, particularly xenopronouns.

Another confounding fact is that there are sometimes exclusive and inclusive words for “I”, kind of. Where there’s a regular way in which duals and plurals can be related to singular pronouns, it’s possible for the two versions of “wit” and “we” to be extrapolated back to “I” and continue to give two forms. That probably isn’t very clear, so I’ll illustrate with fake English pronouns. Suppose we had exclusive and inclusive versions of “we”, such as “wee” and “nee”, and then “yee” for “you”, and the vowel changes in all of them to “oo” in the singular, so “yoo” could be the singular word for “you”. There could then be “woo” and “noo” for the inclusive and exclusive words for “I”. Samoan does this. Its plural inclusive “we” is “mātou” and exclusive “tātou”, the duals are “mā‘ua,” and “tā‘ua” and the singulars “a‘u” and “tā”, although some of these pronouns have variants. Interestingly, you might think that if a number were to collapse into another one to simplify a language, the dual and the plural would merge, but in Samoan the dual and the singular merge instead. “Tā” can mean “I” or “we two”. The exclusive “I” is the usual word, and the inclusive “I” indicates emotional involvement, so for example in “am I going to get one then?”, we use the word “then” to signal emotional involvement, and translating that into Samoan would use the word “tā” and omit “then”. In the closely related Tongan, it’s more connected with modesty and is similar to how posh people use the word “one” in English for “I”.

There’s also a fourth person in some languages, known as the obviative. This is actually not only obviative but obvious, and I used to wonder why it doesn’t happen in English in particular. The obviative contrasts with the proximate, which is the only option in most European languages. It crosses over with the idea of topic prominence, and involves a distinction between more and less important items, so for example, “did you put the food on the table?” where the food is the focus, is “did you put it (PROX) on it (OBV)?” but if it was a table as opposed to a kitchen counter and that was the emphasis, “table” would be the proximate and the food obviative. Presumably this is less important in languages with gender, but as a language without gender it’s seemed odd that we don’t have it. This feeling, unlike the clusivity issue, actually pre-dates any knowledge I had of languages other than English, rather like my daughter’s abortive used of numerical classifiers when she was a toddler, and it makes me wonder how often young children stumble upon features of language absent from their native ones and then reject them as their ability in their first language improves.

English has the distinct and separate reflexive pronouns ending in “-self” and “-selves”. These vary in the third person with dialect, such that “hisself” and “theirselves” is sometimes used in non-standard English consistent with the other pronouns being possessive rather than objective. In many other languages these either don’t exist or the objective forms are used, and in some there is a specific dedicated pronoun for this purpose. In Scandinavian languages this pronoun has become part of the verb and provides them with a mediopassive voice, replacing an older mediopassive found in Gothic. The fact that our own reflexive pronouns are so long means we’re unlikely to develop this way of expressing anything from verbs this way, although the potential used to exist.

The other pronouns also exist. In particular, the English distinction between “who” and “what” is peculiar for our language in that it’s similar to a common vs neuter distinction rather than there being three gender-like forms here. We tend to get confused about “whom” and there seems to be incipient reluctance to say “whose” instead of “of which” or something similar when the referent is inanimate. I said I was going to return to the dual number. In fact this does have some traces in English, and one of these is found in the word “whether”. Nowadays this is used as a conjunction, but it clearly looks a bit like “either” and a bit like “which”, “either” and “neither” being other traces of the dual. “Whether” was previously the dual version of “which”, which has taken over its meaning. There are other remnants too, such as “both” instead of “all”, the slightly vague “couple” and less vague “pair”, and the more contentious “alternative” which can only correctly refer to one of two rather than one of many. Trial pronouns exist in Bislama and Tok Pisin, and also in some Austronesian languages. Lihir also has a paucal number for small numbers of items above three. Paucal seems more obvious than trial and just plain having a plural to me but it’s rare in reality. Sursurunga was thought to have a quadral number but in fact it simply starts to use a different kind of paucal at four and has a “lesser” paucal for two or three, so in fact there seem to be no languages at all, at least right now, with a quadral number.

Vietnamese, I recently discovered through a relative, has a large number of personal pronouns which relate to status and familial relationship, making distinctions which English doesn’t even make with kinship terms. These are all in the second person. Like Japanese and Indonesian/Malay, Vietnamese has formal and informal versions of the first person singular pronoun. The Malay/Indonesian polite word for “I” originally meant “slave”.

There are no languages which don’t have at least two numbers for one or more pronouns. That’s a linguistic universal. However, there are many which manage without what most languages seem to consider vital, including English with its single word for “you”. German shows that it’s possible to get away with a single word for “you”, “she” and “they”, although it doesn’t actually do this most of the time, with “Sie” and “sie”. It probably manages because the verb is inflected differently. Spanish and Portuguese have both adopted noun phrases for polite second person pronouns, namely “Vuestra Merced” and “a senhora”/”o senhor” respectively, and American Spanish rarely uses “tu” and “vosotros”/”vosotras”. Mandarin Chinese substitutes the word for “humble” for the first person singular and “honour” for the second person in polite speech. This is paralleled in the considerably more elaborate Japanese system. In fact it’s been argued that Japanese doesn’t actually have pronouns as such. It’s a topic-prominent language which tends to drop pronouns and nouns often start being used as pronouns which weren’t before. Several other East Asian languages are like this. Indonesian uses the word “tidak” (“thing”) for “it” when “it” refers to something as opposed to in a phrase like “it’s raining”. There’s also the very common phenomenon of pro-drop, which occurs in many European languages whose verbs are sufficiently inflected for the person to be indicated by them. Outside Europe, for instance in Swahili, verbs can be inflected for object, and even indirect object, as well as subject, meaning that a word such as “nitaiosha” means “I will wash it” (in Japanese, incidentally, this same word means “similar company”. There’s a group of unrelated languages whose words can sometimes be identical, Finnish being another.). Pronoun dropping is foreign to all Germanic languages as far as I know except possibly Gothic, although we do have pronoun avoidance. For instance, we sometimes consider “she” and “it” to be rude when referring to human adults.

This brings up the issue of how to avoid pronouns for political reasons, i.e. anti-sexism. English has difficulty with this in regular speech and writing although note form does it. For instance, I might write “Went to the shop” in my diary, although only if the day was particularly boring or I wanted to indicate I was breaking the Sabbath or something. There’s also “Would Madam like some wine?” or “Your Majesty”, “Your Grace”, “Your Worship” and so forth, although these last three include possessive adjectives similar in form to pronouns. The situation in Vietnamese, previously mentioned, seems to be that kinship terms are used instead of the second person.

What’s the minimum number of personal pronouns a language could get away with? Although the answer is obviously “zero”, because it could just use nouns, or each pronoun could have a use as a noun, which happens for example with “Ich” in German, I think the sensible answer is probably two, similarly to “this” and “that”. There’s “this person”/”these people” and “that person or thing”, meaning “you” or “she”/”he”/”it”/”they”. The fact that Gothic doesn’t distinguish between “this” and “that” might indicate that only one is needed, but that language had plenty of other pronouns which might indicate how it coped with that odd deficit. At the other end of the scale, there could be singular, dual, trial, paucal and plural numbers, inclusive and exclusive “we” and “you”, a three-person based system for the third and fourth persons and polite forms for all of them in five genders per person, those being feminine, indefinite gender, neuter, common, masculine and virile (a gender for male persons used in Polish). This technically yields five hundred pronouns, although some of them might make no sense. It can in fact be taken a lot further than that, but five hundred might be enough.

Grammatical And Other Gender

This post might look it belongs on the other blog. The reason it doesn’t will, I hope, become clear as I go on, unless of course I indulge in my usual obscurantist verbosity. I hope I won’t.

Grammatically speaking, English is now a genderless language. This fact leads to confusion because we do seem to have gendered third person singular personal pronouns, and don’t we know it? However, we lack a full grammatical gender system as found in other related European languages. I’m trying to do this post entirely from memory, and unlike the previous one, this one’s supposed to be true, but it may not be because I may not be remembering things correctly. However, from memory, I seem to recall that Armenian lacks gender, which if true is probably because of interaction with the genderless Caucasian languages spoken nearby. Apart from that, Bengali I’ve heard has a gender system practically identical to that of English, i.e. it has gendered third person pronouns which refer mainly to women and men separately plus a neuter pronoun which refers to everything else. Farsi, sometimes called “the English of Asia” due to its simple grammar, lacks gendered pronouns of any kind.

Confusion may result from the assertion that English lacks grammatical gender. So as not to distract you from the rest of what I’m going to say here, I’ll clarify that. In English, we talk about “she”, “it” and “he”, we have indefinite gender singular pronouns “they” and “one” and genderless pronouns such as “this”, “that”, “these” and “those”. We also have a common vs neuter set of pronouns in “who” vs “what”. Other things are also going on. We have an apparently single exception to using “it” for inanimate objects when we talk about ships and boats which is sometimes extended in various ways, for instance to countries and other vehicles. However, we do not have grammatical gender, because a complete grammatical gender system extends way beyond pronouns. The simplest form of gender system in Western European languages which most British English speakers are familiar with is the French system, where nouns are feminine or masculine with no neuter, and this affects not only which pronouns are used with them but also articles and adjectives, including present and past participles. This is about the minimum which can be expected from a real grammatical gender system. English, and Anglo-Norman with which it merged, used to have actual grammatical gender, but this didn’t survive, possibly because the two languages, with different systems, merged. Anglo-Norman French was almost entirely just feminine and masculine, with the lone neuter pronoun «ço» meaning “this”, whereas English had a similar three-gender system to German, with feminine, neuter and masculine. The confusion between referring to nouns in two different gender systems probably led to its demise.

They can be a lot more pervasive than that though. For instance the second and first person pronouns can also be gendered. This happens in the Spanish plural “vosotras” and “vosotros” for “you” in the familiar plural and even in the first person “nosotras”/”nosotros”. Tocharian languages I think even have feminine and masculine first person pronouns. Arabic and Hebrew are part of the Afro-Asiatic language family, which regularly has a feminine and masculine two-gender grammatical system. They have second person feminine and masculine pronouns. In some languages, gender also influences verbs in the sense that there are separate forms for feminine and masculine conjugation using finite verbs rather than just participles. This is about as far as it goes, I think.

Because we live in Europe and are surrounded by languages with gender systems, we tend to assume this is normal for foreign languages, but actually it isn’t. Grammatical gender does occur in hundreds of languages, including most Indo-European and Afro-Asiatic languages, but probably the majority of languages have no gender. The distribution is patchy. Dravidian languages in South India have gender. It’s also sporadically present in Australia, New Guinea (which is not surprising as there are six hundred languages there) and the Americas, with a few in Afrika south of the Sahara. However, the Afrikan gender systems are not related to sex.

Gender as we tend to come across it consists either of feminine and masculine, or feminine, neuter and masculine. The Celtic languages have the first system, as do Western Romance with a few exceptions in Italian, whereas German and the Slavic languages have the second, although Slavic is a bit more complicated than that. However, several European languages have a different kind of gender system which is more like our “who”/”what” system apart from being grammatical rather than semantic. These are the various other Germanic languages: Dutch, Danish, Norwegian and Swedish. I don’t actually know if Frisian has this system, but as it’s the closest continental European language to English it would be interesting to find out. These systems have the common and neuter genders, and there is no feminine or masculine in grammatical terms. In Danish in particular this is very useful, because Danish is a mumbled language whose words can be hard to distinguish. Hence the Danish word for “frog” and “seed” is the same but has common or neuter gender depending on what it means, as is the word “øre”, which either refers to the currency unit or “ear”, again depending on gender. Danish suffers from lacking the tones used in Norwegian and most Swedish which make it easier to distinguish between otherwise similar words, although it does use a glottal stop for a similar purpose. These gender systems, however, have nothing to do with sexes.

Within Europe, several languages completely lack gender. These include Turkish, Hungarian, Finnish, Sami and Basque. They may distinguish by pronouns between animate and inanimate objects, but this has no influence on the rest of the grammar. Caucasian languages, which are also European although we tend not to remember them here in Britain, are also genderless.

There is another phenomenon where gender distinctions are made between different classes of referent but all humans, or animate objects, are of the same gender. These tend to be called “noun classes” rather than genders but they are actually the same kind of system. Swahili has something like nine or ten noun classes. Unlike European genders, they have a very clear system, where for instance artifacts such as knives all have the same class, all animate objects have the same but groups of animates have a different class. Another one is abstract objects. I had a surprising conversation with a Swahili speaker once who didn’t realise this was how Swahili worked, which presumably either means it’s subconscious or they chose to learn each word and how it was pluralised, influenced words around it and so forth, separately. Swahili pronouns are not gendered, so it makes no difference to the one thing English speakers would expect it to.

Mandarin Chinese and Malay/Indonesian (which are more or less the same language) have no gender as such, but they do have a system of counting which classifies the objects counted in a similar way. In the same way as we say “fifteen head of cattle”, Indonesian counts animals by tails, “ekor”, and it and Mandarin have obligatory words inserted before the noun. Our daughter actually used to do this in English when she was first learning to speak in full sentences, referring to certain objects as “loaves” when she counted them but not others. Japanese kind of has first person singular gendered pronouns because when they do use their words for “I” and “me”, women tend to use different words than men. This is strange because there is no other gender at all in Japanese.

Turning now to the history of grammatical gender in English, I will begin with prehistory. Proto-Indo-European itself was spoken before the invention of writing, but can be reconstructed. Hittite and the other Anatolian languages, spoken at the start of the Iron Age, are a kind of sister group to all later recorded languages in the family and are therefore a kind of fossilised stage in the development of the original language which indicates the process whereby gender systems emerged. It basically works like this. In Hittite there are two genders, animate and inanimate. This seems quite logical to me, and is similar to the “common”/”neuter” system found in Scandinavian languages and Dutch. These did function as genders, changing the form of adjectives associated with them and being inflected differently in the nine cases, but they are only somewhat like the systems found in the likes of Sanskrit, Greek and Latin. What is thought to have happened is that there were three classes of referent approached differently in terms of grammar. There were most inanimate objects, and we would refer to each of them as “it”. Then there were two other classes of object, one consisting of agents and the other of adjectives. Hence a spear might be in the agent class, because it does something, whereas a flower might be in the adjectival class because it’s beautiful, i.e. mainly conceived of by an attribute. Since men were perceived as the “doers”, they were masculine, and because women were perceived as being rather than doing, e.g. being beautiful, they were feminine. It’s important to note, though, that it isn’t a case of a whole class of items being considered feminine in the human sense, or another whole class being considered masculine. It’s completely the other way round. There is a class of objects thought of as what they are, including women, and another class of objects thought of as what they do, and this includes men. Sex is secondary. Then there’s a third class which falls into neither category. That’s how it happened.

This was also only the initial situation. Words change their meaning over time, people think words which rhyme have the same gender and so on. All Latin names for trees are second declension feminine nouns even though they end in “-us” like all the second declension masculine nouns. “Manus” is also feminine. Due to the similarity, in Italian and related language, all these nouns have now become masculine. There has also been confusion between first declension neuter plurals ending in “-a” and feminine nouns, so for example the plural of “opus” is “opera” , but the latter is now a singular feminine noun.

Back to English though. Proto-Germanic had three genders: feminine, neuter and masculine. Originally, the words for “they” were also gendered. Gothic has “ijos”/”eis”/”ija” for “they” in the plural, with separate forms for feminine, neuter and masculine. Icelandic has the same distinction: “þær” “þau” “þeir”. Icelandic is important to English pronouns because our own third person plural pronoun, “they”, is inherited not from West Germanic but Old Norse, which is practically the same language. In Old English, by the time it was written down at least, made no distinction between its plural words for “they”. The only survival from this word is today’s “’em” for “them”, which is colloquial, but the word was “heo”/”hie” in the nominative and accusative, “him”/”heom” in the dative and instrumental and “hira” in the genitive. This compares to the singular feminine “heo”/”hie” in nominative and accusative, “hire” in the genitive, dative and instrumental. The neuter and masculine are the same as each other in the dative and instrumental and in the genitive, i.e. “him” and “his”, but otherwise differ, being “hit” and “he” in the nominative and “hit” and “hine” in the accusative. This mixture of somewhat confused forms is typical of pronouns in West Germanic: it’s seen today in the German “sie” and “Sie” for “she”, “they” and the polite form of “you”, and in “ihr” and “Ihr” for the informal plural “you” in the nominative, “her” in the genitive and “their” and polite “your”, again in the genitive. I haven’t conversed in Old English as much as in German, but I’ve never encountered any problems with the fact that these are homonyms. Notably, their word for “they” is often the same as their word for “you” and it doesn’t cause confusion.

The reason I’ve written this out rather than used a table is to emphasise the detail of what’s going on. In some circumstances, the word for “they” is the same as the one for “she” and in others it’s the same as the words for “it” and “he”. There is therefore, just as in German today, a lot of ambiguity here. As with German, this is often resolved by the form of the verb, because it would be plural when the pronoun is plural and singular along with the pronoun. In the meantime, “his” and “him” could be neuter or masculine. The ambiguity might, of course, be the reason we ended up with an Icelandic word for “they”, but it’s also interesting that we don’t say “him” or “‘im” even informally to replace “them”, but instead use a form derived from the Old English dative and instrumental, in both objective forms.

These, though, are just the pronouns, and in Old English there was a complete grammatical gender system. Each noun was feminine, neuter or masculine. Bear in mind that this was grammatical gender, not gender as the word is often used in English. The gender of a noun, as in many other language, was determined by its ending, including complete words. For instance, there’s a street in Canterbury called Burgate, from a feminine word for “city”, “burg”, and the neuter word for “gate”, “geat”. The whole word “burg-geat” is neuter, because the last word in the compound is neuter. Every noun that ends with “-a” is masculine, compared to the Latin tendency for such nouns to be feminine (but not always, e.g. “agricola”, “nauta”). There were various suffixes which conferred gender reliably such as “-dom”, as in “wisdom”, “-had”, as in “cildhad” (“childhood”) and “-scipe”, as in “friendship”, all masculine, and the feminine “-nes”, “-o”, “-ræde” and “-ung”: “rihtwisnes” – “righteousness”; “bieldo” – “boldness”; “hatræde” – “hatred” (that might be a spelling mistake on my part); “scotung” – “shooting”. One word for “man”, “mann” is masculine, as is “wer”, also meaning “man” as in “werewolf”. However, the main word for “woman” is “wif”, i.e. “wife”, and this word is neuter. Moreover, the ancestor of our current “woman” was “wifmann”, which is masculine. “Cild” and “bearn”, both words for children, are neuter, as is “mægden” (“maiden”). Hence if you started a sentence talking about a “wifman”, you would then refer to that person as “he” throughout it. Similar things happen today in Gàidhlig incidentally, so it still happens in Britain. I am personally accustomed to saying “it” when referring to a child or baby. The takeaway from this is that the whole time Old English was spoken there was only a loose anchor between pronouns and what we might call biological sex, or gender in the social sense, and likewise between grammatical and social gender.

Going back to the Old English pronouns, it’s notable that the forms of the singular third person have changed a fair bit and the plural has practically disappeared. The situation is complicated by the fact that the dominant dialect of English over most of the period before the Norman Conquest was West Saxon, spoken in Wessex, whereas the English in which this is written is descended from the Mercian dialect, spoken in the English Midlands. In particular, this means that the diphthongs written as “eo” and “ea” are not the direct ancestors of any modern English sounds. Hence the “heo” and “heom” above might be said not to survive, although related words in other dialects did. The distinctions between the likes of final A, E and U were lost, replaced by the schwa “murmur” vowel. This meant in turn that any grammatical distinction made through these vowels was also lost, and since these, among other things, were responsible for gender distinctions, these went too. Hence although different pronouns might’ve been used to refer to different nouns, they weren’t accompanied by as complex a gender system as previously. All that was left was an occasional “-e” appended to adjectives in the plural and after “þe”, as in “þe olde worlde”. Gender would still have shown up with the various demonstrative pronouns, although even here they were more inflected in the South. This left the personal pronouns, which were also changing.

Of course “he” still existed, as it does today. “It” emerged due to a dropped H. “Hie” survived at least into Chaucer’s time as “hi” and the genitive “hir”. “Heo”, the former feminine pronoun, turned into a confusing plethora of different words, including “he” and “ho”, and “hi” was accompanied by “he” as well as the new Scandinavian forms. In other words, it became entirely feasible for the word “he” to mean what we mean by “she” or “they”. This is the origin of generic “he”. There was pressure to resolve this situation, which was achieved by using the Scandinavian pronoun in the plural most of the time and by adapting the feminine “that”, which was “seo”, except of course that that diphthong had been lost, leading finally to “she”. So there was an apparent felt need to use a feminine pronoun at this point. Interestingly, and probably coincidentally, everyone was basically wearing dresses at this point too, the only gender difference being the heights of the waists. It’s hard to imagine how this situation would have arisen if the Bible had been interpreted the way it often is today, but that’s another conversation. “She” prevailed in about 1300, but the use of generic “he” persisted all the way into my lifetime and is even sometimes the source of contention today in 2023. The use of generic “he” is a good illustration of how the history of a word may not have any bearing on how it’s taken today. Nowadays, using generic “he” just is sexist, regardless of its history of originally including the feminine. I may be very attached to the history of language but I still recognise that it has its place. It does, for example, illustrate that the ability to refer to social gender at that point was considered vital, and this is not trivial as there were languages in similar situations where it wasn’t, so it says something about English society that this happened.

A more distinct animate third person singular pronoun was “ha”. This was also gender-neutral, but resembles neither “she”, “he” or “they” (“hi”) very closely. It persisted into the last century in some West Country speech and was adopted by Ursula Le Guin for the screenplay of her novel ‘The Left Hand Of Darkness’. It’s used sparingly today as a neutral pronoun, but is probably too exotic to catch on. However, it does date back to the Middle Ages.

Now for the question of apparently plural pronouns being used in the singular.

This occurs, for example, in Urdu, where married women tend to say “ham” (we) rather than “main” (I) because they’re used to referring to their whole household, and there’s the “editorial ‘we'” and the “royal ‘we'”. The former is when someone acts as a spokesperson and the latter is used by royalty in official proclamations. It also appears to crop up in the Bible but this may be a disguise for the possibility of these passages being polytheistic.

English is unusual in not using the singular “thou”, and in this respect it follows Anglo-Norman, which stopped using «tu» in favour of «vos», so I wonder if there’s a connection. The former English use of “ye” and “you” followed the French usage of these pronouns, the latter now being «vous», in that the plural was used for both the plural itself and formally, whereas the latter was informal and only ever singular. It’s used slightly differently in the King James Version of the Bible, simply to translate the singular and plural pronouns, making it look like “thou” is formal due to God being referred to in that way. This may have led to the erosion of the distinction in English, but I’m just guessing. It’s clearly in common use in 1611 CE, and also in the works of Shakespeare to 1613 which doesn’t get us (“me”?) much further. What seems to have happened is that during the seventeenth century, using “you” was thought polite and people were expected to be polite all the time. “Y’all” and “youse” are clearly attempts to address this, as is “you lot”, although that is definitely not polite. Quakers were well known to thou well past the time when standard English stopped, although they don’t do it any more and so it’s now only used dialectally. It’s also true that you’d probably have to know someone pretty well before you called them “thee”.

Bearing in mind the time scale, this finally brings me to “they”, which is enjoying a moment. “They” hadn’t planted itself in our language at the time of Chaucer, or at least the London dialect thereof. Being a Scandinavian word, it unsurprisingly moved from North to South. It’s used once in the text ‘Genesis And Exodus’, written in 1250 in Norfolk. Before that, the ‘Ormulum’, written in the previous century in the East Midlands, uses it, but since this area, where I happen to be sitting right now, had fairly recently been Danish, it’s not surprising it was present. The language of the Ormulum is idiosyncratic, using a spelling system seen nowhere else, and it’s also quite badly written. Also in ‘Genesis And Exodus’, “it” is used as a plural, and the word “his”, presumably because it has an S at the end, is too. All of this is going on while the English language is eclipsed by Norman French, so it was kind of decaying at the time. Chaucer does sometimes use “they” but not “their” or “them”. He died in 1400. There are also the reflexive pronouns, which nowadays are the likes of “herself”, “myself” and so on. Back then, they were used interchangeably with object pronouns, and the plural “-selves” forms didn’t exist. Therefore, “themself” crops up quite early, though not necessarily with a singular meaning.

The plural “they” appeared about a century before singular “they”. Grammatical gender was lost during the thirteenth century, just as the word “they” was being adopted into English. This could benefit from some grammatical context. American English uses generic “he” after “one”, as in “one does what he pleases”, whereas British English doesn’t: “one does what one pleases”. Singular “they” has tended to crop up in rather similar circumstances: “Had the Doctor been contented to take my dining tables as any body in their senses would have done …” – Jane Austen, ‘Mansfield Park’, 1814. The use of the English language before the eighteenth century was not strongly prescribed, but from that time on, such uses of “they” tended to be frowned upon on the grounds that it came across as plural, and “he or she” on the grounds that it was clumsy. Three important points come to mind here. Firstly, it’s actually three centuries older than singular “you”. Secondly, the use of generic “he” rather than “they” seems to be more American than British. Now I’m quite a fan of American English. For instance, I like the precision of “gotten” as a past participle better than “got”. Thirdly, Jane Austen used it, whose language is quite classical, focussed on sentence structure and balanced. It is of course possible that she was deliberately avoiding generic “he” because she mainly writes about female characters.

It is entirely standard usage to say something like “when the interview candidate comes in, make sure they have the right chair”, and I think in general that would go without comment or even being consciously noticed by most people. It’s a usage before meeting or otherwise encountering the person in question, and it may be the other singular usage, which seems to be newer, which bothers people. However, although I do think there are problems with the use of singular “they” which I’m going to outline later, I do think there’s a new usage, which is however difficult to define. It’s along the lines of assumed gender. Singular they is more often used in advance when someone’s gender is not known to be feminine or masculine. It’s now also used in arrears in this situation. That is, people now either claim the usage as a reported pronoun or it could potentially be used because appearances are sometimes deceptive. If a fool’s tiger and a real tiger were to have names with different grammatical genders, it might be advisable to use singular “they” if one didn’t know what one was seeing. It is quite possibly a question of evidence for gender: what was previously considered a sufficient condition for being gendered in a particular way is no longer, and only honest reporting of one’s pronoun could work. That’s probably the idea anyway.

It may be partly a question of semantic drift. I’m fond of saying that the words “silly”, “nice” and “gay” have changed their meanings dramatically, and in fact they overlap each other sometimes. It used to be common for people to object to the use of the word “gay” to mean “male homosexual”, but it sometimes seemed that they weren’t really deprived of a word and they could’ve said “gaudy”, “merry”, “happy” or many other words in its place. However, I don’t think this is all.

The English language is not very inflected. The present tense of regular verbs has only one variant, and sometimes not even that, namely “-s” for the third person singular. Hence if one does use singular “they”, maybe one could say “they is” or “they was”, but this is grating. There could potentially be a lack of important information. Other approaches can be taken though, such as wording things in the passive, using a genuinely plural “they”.

In conclusion then, grammatical gender is not gender, although it has been said that, for example, speakers of a language whose word for “bridge” is feminine will tend to use adjectives such as “elegant” to describe it, whereas when they speak a language whose word is masculine tend to describe them as “sturdy” and so forth, so it does seem to have a psychological effect. Singular “they” is older than singular “you”, and it may lose information to use it, but we’ve survived singular “you”. I actually prefer “it” because it’s a great leveller, but I’m also sure it’s very unpopular.