Tag Archives: germanic

The relative chronology of Grimm’s Law and Verner’s Law, part 1: aspiration in the Germanic languages.

Grimm’s Law and Verner’s Law are possibly the two most famous sound laws in historical linguistics. Despite this, there are some aspects of these two laws which we know little about. One of these is the question of the relative chronology of the sound changes described by these laws. That is: which came first? The sound changes described by Grimm’s Law, or the sound changes described by Verner’s Law? Handbooks, such as Ringe (2006), tend to ascribe to the view that those described by Grimm’s Law came first, and those described by Verner’s Law came second. But as I’m going to attempt to show, this is not a completely well-established fact.

Now, strictly speaking, Grimm’s Law and Verner’s Law describe correspondances between the sounds of Proto-Indo-European (PIE) and Proto-Germanic (PGmc); the actual sound changes that have resulted in these correspondances are another matter. The correspondances are very well-established; there is little disagreement over them. So one might well say that the question posed here is uninteresting, because we know which PGmc sounds reflect which PIE sounds in which positions, and that’s all we need to know. This is true to some extent, but I do think it is interesting in its own right to know more about the relative chronology of the sound changes that turned PIE into PGmc. Besides, our understanding of what a sound change must have been, in phonetic terms, can be affected by our understanding of its relative chronology, and this understanding may help us to understand the nature of other sound changes, or of the phonology of the language at an earlier or later date. More knowledge is usually a good thing, after all. (But it doesn’t surprise me that I can’t find much literature dealing with this issue specifically.)

With that said, let’s begin by reminding ourselves of the correspondences described by Grimm’s Law, which are listed in the table below.

Proto-Indo-European Proto-Germanic Example
*p *f PIE *pl̥h₁nós ‘full’ (cf. Skt pūrṇás, Lith. pìlnas) ↣ PGmc *fullaz (with the -az ending generalised from thematic nominals without stress on the ending) (cf. Goth. fulls, OE full [> NE full])
*t PIE *tréyes ‘three’ (cf. Skt trayaḥ, Grk treîs) > PGmc *þrīz (cf. Goth. þreis, OE þrī [↣ NE three])
*ḱ *h PIE *swéḱuros ‘father-in-law’ (cf. Skt śvaśuraḥ, OCS svekrŭ) > PGmc *swehuraz (cf. OE swēor, OHG swehur)
*k *h PIE *kóryos ‘army’ (cf. dialectal Lith. kãrias ‘army’, OIr. cuire ‘troop’) > PGmc *harjaz (cf. Goth. harjis, OE here)
*kʷ *hʷ PIE *ákʷah₂ ‘running water’ (cf. Lat. aqua ‘water’) > PGmc *ahʷō ‘river’ (cf. Goth. aƕa, OE ēa)
*b *p post-PIE *gʰreyb- ‘grab’ (cf. dialectal Lith. greĩbti [infinitive in -ti]) ↣ PGmc *grīpaną (infinitive in -aną) (cf. Goth. greipan, OE grīpan [> NE grip])
*d *t PIE *dóru ‘tree’ (cf. Skt dā́ru, Gk dóru ‘wood’), gen. sg. *dréws (cf. Skt drós) ↣ PGmc *trewą (with the neuter a-stem ending ) (cf.
*ǵ *k PIE *h₂áǵros ‘pasture’ (cf. Skt ájras ‘field’, Lat. ager) > PGmc *akraz (cf. Goth. akrs ‘field’, OE æcer ‘field’)
*g *k PIE *yugóm ‘yoke’ (cf. Skt yugám, Lat. iugum) > PGmc *juką (cf. Goth. juk, OE ġeoc)
*gʷ *kʷ PIE *gʷih₃wós ‘alive’ (cf. Skt jīváḥ, Gk zōós) > *kʷikʷaz (cf. ON kvikr, OE cwic)
*bʰ *b PIE *bʰéreti ‘(s)he is carrying’ (cf. Skt bhárati, Lat. fert) > PGmc *beraną (infinitive in -aną) (cf. Goth. baíran, OE beran)
*dʰ *d PIE *dʰédʰēm ‘I was putting’ (cf. Skt ádadhām [with the augment á-]) > PGmc *dedǭ ‘(s)he did’ (cf. OS deda, OHG teta)
*ǵʰ *g PIE *ǵʰáns ‘goose’ (cf. Gk khḗn, Lith. žąsìs [with the i-stem ending -is]) > PGmc *gans
*gʰ *g PIE *gʰóstis ‘stranger’ (cf. Lat. hostis ‘enemy’, OCS gostĭ ‘guest’) > PGmc *gastiz ‘guest’ (cf. Goth. gasts, OE ġiest)
*gʷʰ *gʷ PIE *sengʷʰ- ‘chant’ (cf. collective *songʷʰáh₂ > Gk omphḗ ‘voice of the gods’) > PGmc infinitive singʷaną ‘to sing’

Basically, the PIE voiceless unaspirated stops become fricatives, the PIE voiced unaspirated stops lose their voice, and the PIE voiced aspirated stops lose their aspiration. (But this is not quite a complete description of what happened, as we will see.)

The correspondances described by Grimm’s Law do not hold in every position. One position which they do not hold in is position after a voiceless obstruent. In this position, PIE voiceless unaspirated stops do not become fricatives in PGmc, and thus end up being reflected as the same kind of sound that the PIE voiced unaspirated stops are reflected as in other positions. Here is a full list of the clusters affected by this change, with examples.

Proto-Indo-European Proto-Germanic Example
*sp *sp PIE *spŕ̥dhs ‘contest’ (c.f. Skt spṛdh > PGmc *spurdz ‘racecourse’ (cf. Goth. spaúrds)
*st *st PIE *gʰóstis ‘stranger’ (cf. Lat. hostis ‘enemy’, OCS gostĭ ‘guest’) > PGmc *gastiz ‘guest’ (cf. Goth. gasts, OE ġiest)
*sḱ *sk PIE *sḱinédsti ‘(s)he cuts (it) off’ (cf. Skt chinátti), aor. sbjv. *skéydeti ↣ PGmc infinitive skītaną ‘to defecate’ (cf. ON skíta, OE scītan)
*sk *sk PIE *skabʰeti ‘(s)he is scratching’ (cf. Lat. scabit) > PGmc *skabidi or *skabiþi (cf. Goth. skabiþ, OE scæfþ)
*skʷ *skʷ (no examples that I know of, but this outcome can be assumed on the basis of the others)
*pt *ft PIE *kh₂ptós ‘grabbed’ (cf. Lat. captus ‘caught’) > PGmc *haftaz (cf. OE hæft, OHG haft)
*ḱt *ht PIE *oḱtṓw ‘eight’ (cf. Skt aṩṭā́u, Lat. octō) > PGmc *ahtōu (cf. Goth. ahtau, OE eahta)
*kt *ht PIE *mogʰ- ‘be able to’ (cf. Skt maghám ‘possessions’ [a-stem pl. in -ám], OCS mošti ‘I can’ [infinitive in -ti]) → nominal *mógʰtis > PGmc mahtiz ‘power’
*kʷt *ht PIE *nókʷts ‘night’ (cf. Gk núx, Lat. nox) > PGmc *nahts (cf. Goth. nahts, OHG naht)

Now, here’s an interesting observation: in English, there is a rule that voiceless stops (which are, in English, directly inherited from Proto-Germanic for the most part) are aspirated except after another voiceless obstruent: hence in my dialect of English tale is pronounced [ˈtʰejəɫ] (in my dialect, anyway) while stale is pronounced [ˈstejəɫ]. There may be other environments where there is no aspiration, depending on dialect and perhaps individual variation (for example, word-final voiceless stops can be aspirated, glottalised, unreleased or none of these things; and my own dialect tends to fricativise them, although this is one of its more idiosyncratic features). Also, it is possible for there to be different degrees of aspiration, which complicates matters further. But there is definitely no aspiration after a voiceless obstruent, and there is definitely a maximal level of aspiration when a stop is word-initial, or in the onset of a stressed syllable (as in attack).

The same rule is observable in most of the other Germanic languages. The only exception I know of is Dutch, in which voiceless stops are attributable in all positions, but this may be attributable to the influence of French. The case of German is particularly interesting, because in German, the stop t does not reflect Proto-Germanic *t; that phoneme became either z (the affricate /t͡s/) or s, depending on its position, in German due to the High German consonant shift. German t instead reflects Proto-Germanic *d, which filled the gap in the consonant system left by the loss of *t by losing its voice. Yet German t obeys the aspiration rule just like the other plosives. It is of course possible that the aspiration rule is simply something that came into effect after the separation of the Germanic languages after the devoicing of *d in the High German dialects. But in that case, it would have had to come into affect in all of the non-Dutch Germanic languages independently. Furthermore, the development of the PGmc voiceless stops in the High German consonant shift suggests that these voiceless stops were aspirated at the time of the shift, because as far as I know, the development of voiceless stops into affricates, when not motivated by palatalisation, tends to occur only when they are aspirated. After all, affrication under these circumstances can be explained as assimilation of the phonetic [h] that follows the release of voiceless aspirated stops to the place of articulation of the preceding stop; I know of no reason why unaspirated stops could be expected to turn into affricates. Lenition alone cannot account for affrication, because affricates involve just as much stricture, during their initial stop articulation, as stops.

For these reasons, I think it is more likely that this aspiration rule was inherited from Proto-Germanic into all of the Germanic languages, and that it persisted in German after the High German consonant shift, applying to the new instances of t produced by this shift. It is entirely possible for phonological rules to persist in this way. For example, Siever’s Law, the phonological rule that caused underlyingly non-syllabic PIE sonorants to become syllabic after heavy syllables, persisted into Proto-Germanic, as can be seen from the example of PIE *wr̥ǵjéti ‘(s)he is working’ (cf. Av. vərəziieiti) > *wurkijiþi > PGmc *wurkīþi (c.f. Goth. waúrkeiþ, OE wyrcþ).

Now, if you accept that the aspiration rule could have persisted in applying after the High German consonant shift, it’s no stretch to suppose that the aspiration rule took effect before the sound changes described by Grimm’s Law occured, and it persisted in applying to the new voiceless stops produced by these changes. Why would we want to suppose this? Because it allows us to neatly explain the fact that the PIE voiceless stops did not become fricatives after voiceless obstruents. Position after voiceless obstruents is exactly the position where these voiceless stops did not become aspirated by the aspiration rule. So if the aspiration rule did take effect before the sound changes described by Grimm’s Law, those sound changes applied precisely to the aspirated voiceless stops, in all positions, and not the unaspirated voiceless stops. And fricativization of voiceless aspirated stops but not voiceless unaspirated stops is well-attested from languages such as Greek (consider: theós = classical [tʰeós], modern [θɛˈɔs], treîs = classical [tré͡es], modern [ˈtris]).

Readers (if I have any?) might remember that I already proposed this scenario in an earlier post. But I don’t have any formal qualifications in linguistics (yet!), so I can’t be regarded as a reliable source. However, I did find a reassuring paper by Iverson & Salamon (1995) which proposes the same scenario. What’s more, they also provide convincing phonetic motivations for why it was the voiceless aspirated stops that became fricatives, rather than the voiceless unaspirated stops or both kinds of stops, and for why voiceless stops after voiceless obstruents failed to become aspirated in the first place.

In phonetic terms, voiceless aspirated stops are distinguished from voiceless unaspirated stops by the fact that the open state of the glottis which is required in order to produce a voiceless sound persists for a short period after the release of a voiceless aspirated stop (this might be achieved by closing the glottis more slowly, beginning with a wider glottal opening in the first place, or a combination of the two). This results in the production of a phonetic [h] sound ([h] being the sound obtained when air passes through the open glottis and out of the mouth without being obstructed in the oral tract), although this [h] sound is considered part of the aspirated stop, in phonological terms. (Languages which have a /h/ phoneme as well as voiceless aspirated stops may distinguish phonemic /h/ by its longer duration; compare the English near-minimal pairs deckhand and decad.) Hence voiceless aspirated stops endure for some time after their release. Voiceless unaspirated stops, on the other hand, do not; after their release, the glottis shifts almost immediately to the state required for the production of the next sound (or comes to rest, if a pause follows). Now, if we assume that there is a tendency for stop phonemes to have similar durations, it follows that we should expect voiceless aspirated stops to have a shorter duration up to the release, that is of the period of obstruction, than voiceless unaspirated stops. And this has been backed up by empirical observations. Because the period of obstruction is shorter in voiceless aspirated stops, there is a greater tendency for the obstruction to be weakened, for whatever reason (e.g. a natural tendency towards weakening of shorter sounds, or assimilation to neighbouring sounds whose production involves less obstruction in the oral tract). That is why the obstruction tends to be weakened from the complete closure required for a stop to mere close approximation, which results in a fricative sound.

As for the question of why the PIE voiceless unaspirated stops did not become aspirated after voiceless obstruents in pre-PGmc, Iverson & Salamon answer this by proposing that the [+spread glottis] feature (i.e. the feature of extending the period of glottal opening by closing the glottis more slowly, or beginning with a wider glottal opening in the first place, or a combination of both of these things) is shared between the constituent consonants in a cluster of two consonants in which the first is an obstruent. That means that the extended period of voicelessness, which normally manifests as the phonetic [h] sound that follows an aspirated stop, is absorbed by the second constituent consonant in the cluster. Clusters like /st/ start off being pronounced with a glottis which is as widely spread as it is at the start of an aspirated stop, and over the course of the cluster the glottis closes just as slowly; by the time the end of the cluster is reached, the glottis is closed enough that there is no discernable [h] sound at the end.

This is not the only phenomenon observable in the Germanic languages that can be explained by this proposal that the [+spread glottis] feature is shared in biconsonantal obstruent-initial clusters. In English, for example, sonorant consonants after tautosyllabic voiceless obstruents are, generally, devoiced. But they are not devoiced after tautosyllabic /s/ + voiceless stop clusters (e.g. in /spl/ and /spr/). If this devoicing is just a matter of perseverant assimilation, this is difficult to explain. But if the devoicing is the effect of the extended period of voicelessness following a voiceless aspirated stop, it is exactly what we would expect. Iverson & Salamon don’t mention if the same pattern is found in other Germanic languages, but we would expect it to be found in all of them except Dutch.

So, that’s the first exception to Grimm’s Law. The second exception is the one described by Verner’s Law. But this seems like a good point to pause for now; I’ll cover Verner’s Law, and its relative chronology, in another post. (This post hasn’t been wholly unrelated to that topic; the observation that PIE voiceless unaspirated stops probably became aspirated in most positions before the sound changes described by Grimm’s Law is going to be relevant.)


Why are German adjectives so weird?

As in many other languages, German adjectives take different endings depending on the case, number and gender of the nouns that they are associated with. For example, in the two noun phrases below, the adjective schwarz ‘black’ appears in two different forms due to the fact that the nouns Hund ‘dog’ and Katze have different genders (masculine and feminine, respectively).

ein schwarzer Hund
a black dog

eine schwarze Katze
a black cat

Although it can be difficult for speakers of languages that have invariant adjectives, like English, to get used to this feature of German, it’s not a particularly remarkable one. It’s an example of the phenomenon known as agreement, which is very common in natural languages. Indeed, agreement occurs in English, although it is less pervasive than in German, occuring consistently only with finite verbs in the present tense, which take the ending -s when their subject is third-person and singular; for example, the verb look appears in two different forms in the two sentences below, because in one of them its subject is third-person or singular, while in the other its subject is third-person and plural.

It looks good.

They look good.

But there is one feature of German adjectives which is decidedly odd. For each particular combination of case, number and gender, there are actually two different adjective endings associated with that combination. One of them is called the strong ending, and the other is called the weak ending. Now, it’s not unusual for there to be multiple sets of adjectival endings in a language, with different adjectives taking endings from different sets. For example, in Latin there are five different sets of endings (or declensions), and in order to put the correct ending on an adjective you have to know which declension it takes. In the two noun phrases below, the adjectives imperiōsus ‘powerful’ and juvenis ‘young’ take different endings, even though they are both agreeing with a masculine noun in the nominative plural, because imperiōsus takes the first/second declension1, while juvenis takes the third declension.

imperiōsī virī
powerful men

juvenēs virī
young men

But the weird thing about German adjectives is that every adjective takes both sets of endings depending on the context in which it occurs. This is quite different from the situation in Latin; in Latin a given adjective always takes the same set of endings. For example, look at what happens if we alter the two German noun phrases given above slightly, so that they begin with the definite article (‘the’) rather than the indefinite article (‘a’):

der schwarze Hund
the black dog

die schwarze Katze
the black cat

The adjective schwarz takes the weak endings, rather than the strong endings, in these two noun phrases, and we can see the difference in the first one, because the strong masculine nominative singular ending is -er, but the weak masculine nominative singular ending is -e. There is no change to the form of schwarz in the second noun phrase, but that’s just because the strong and weak feminine nominative singular endings happen to be identically -e.

The question that I’m going to explore in this post is this: why does this weird system, where there are two different sets of adjectival endings, exist? What purpose does it serve, if any?

First, I should explain how the system works in more detail. The complete sets of strong and weak German adjectival endings are given in the following two tables.

German strong adjectival endings
Singular Plural
Masculine Feminine Neuter
Nominative -er -e -es -e
Accusative -en -e -es -e
Genitive -en -er -en -er
Dative -em -er -em -en
German weak adjectival endings
Singular Plural
Masculine Feminine Neuter
Nominative -e -e -e -en
Accusative -en -e -e -en
Genitive -en -en -en -en
Dative -en -en -en -en

One thing you can see from these tables is that that the weak endings exhibit a lot of syncretism (the use of the same ending for multiple case-number-gender combinations). In fact, the weak endings only take two distinct forms, -e and -en. They are therefore of little use if you want to use them to determine the case, number and/or gender of the noun associated with the adjective. On the other hand, the strong adjective endings, though not free of syncretism, are sufficiently varied that the ending of a strong adjective, alone, gives you a pretty good indication of the case, gender and number of the associated noun. In other words, the strong adjective endings mark case, gender and number effectively, while the weak endings don’t mark case, gender and number effectively. If the purpose of agreement is to make distinctions of case, number and gender more noticeable, then the strong endings are a lot more functional than the weak ones. This observation will be important later on.

But first, there’s another essential piece of information we need to know: when are the strong endings used, and when are the weak endings used? Well, first I should note that adjectives may not take any endings at all. German adjectives only take endings when they are attributive, i.e. when they come directly before the noun they are associated with. Adjectives can also be linked to a noun indirectly, most commonly via the verb wesen ‘to be’, in which case they are said to be predicative. For example, in the following sentence, schwarz is predicative and therefore does not have the weak nominative feminine singular ending -e as it would if it were attributive.

Die Katze ist schwarz.
The cat is black.

One way of thinking about this is that predicative adjectives are actually being used as adverbs; schwarz in the above sentence is associated to the verb ist, rather than to the subject of this verb, die Katze, and it accordingly takes no endings, as usual for adverbs.

When adjectives are attributive, they often come after another word which is called a determiner. Determiners are rather like adjectives in that they are associated with nouns and come directly before them; the main differences are that when a determiner and an adjective are both associated to a noun, the determiner goes before the adjective, and determiners are never predicative, although the distinction is not entirely clear-cut. The most common determiners are the two articles, der ‘the’ (the definite article) and ein ‘a’ (the indefinite article). In German (and English), every singular noun has to be associated to a determiner; the only time nouns do not have determiners in front of them are when they are plural (or if they are abstract nouns or mass nouns, which are unmarked for number) and indefinite, since ein was originally the numeral 1, so it can only be associated with singular nouns).2

It is the determiner that comes before an adjective that determines which set of endings it will take. As you already saw in the example above, adjectives that come after the definite article der take weak endings, and adjectives that come after the indefinite article ein take strong endings, although there is actually a catch in the latter case: when the noun associated to an adjective that comes after the indefinite article is in the dative or genitive case, rather than the nominative or accusative case, the adjective takes weak endings rather than strong endings. Of course, the genitive and dative weak endings all take the same form (-en), so this is really only a minor amendment to the general pattern that adjectives after ein take strong endings. But the set of endings that adjectives after ein take is sometimes treated as a separate set of endings called the mixed endings; the table is shown below for clarity.

German mixed adjectival endings
Singular Plural
Masculine Feminine Neuter
Nominative -er -e -es -e
Accusative -en -e -es -e
Genitive -en -en -en -en
Dative -en -en -en -en

There are other determiners besides the definite and indefinite articles. Some of them are der-like, in that adjectives that come after them take weak endings. These include dieser ‘this’ (commonly also ‘that’ in modern usage), jener ‘that’ (rare in modern usage), welcher ‘which’, solcher ‘such’, jeder ‘every’ and mancher ‘many’. Others are ein-like, in that adjectives that come after them take mixed endings. Apart from ein, the only other ein-like determiners are kein ‘no’ and the possessive determiners derived from personal pronouns: mein ‘my’, dein ‘your (singular, informal)’, sein ‘his, its’, ihr ‘her, their’, unser ‘our’, euer ‘your (plural, informal)’ and Ihr ‘your (formal)’. Both the der-like and ein-like adjectives are declinable: they take different endings depending on the case, number and gender of the nouns they are associated with. Interestingly, der and the der-like determiners all take one set of endings (although the endings take slightly different forms when added to the stem d- of der, because the stem d- does not constitute a whole syllable by itself and hence endings that come after it are stressed), while ein and the ein-like determiners all take a different set of endings. These endings are shown in the two tables below.

German der-like determiner endings
Singular Plural
Masculine Feminine Neuter
Nominative -er -e (stressed: -ie) -es (stressed: -as) -e (stressed: -ie)
Accusative -en -e (stressed: -ie) -es (stressed: -as) -e (stressed: -ie)
Genitive -es -er -es -er
Dative -em -er -em -en
German ein-like determiner endings
Singular Plural
Masculine Feminine Neuter
Nominative -∅ -e -∅ -e
Accusative -en -e -∅ -e
Genitive -es -er -es -er
Dative -em -er -em -en

There are also determiners which are indeclinable (i.e. they take no endings). Note that not all of these would normally be regarded as determiners, but it is convenient to think of them as determiners for the purposes of this post. This group contains quite a few different kinds of word:

  • Numerals greater than two, although ein ‘one’ is not included because it is the same word that is used as the indefinite article and hence takes the ein-like endings. German is thus unlike English, in which a and one are separate words in every respect, although it is possible to tell via intonation when ein is being used as the numeral 1 or as the indefinite article.
  • Quantifiers derived from noun phrases such as ein bisschen ‘a little’ and ein paar ‘a few’.
  • The comparative particles, mehr ‘more’ and weniger ‘less’.
  • Proper nouns in the genitive case such as Jesu ‘Jesus”, Vatis ‘dad’s’ or Muttis ‘mum’s’, which, unlike other nouns in the genitive case, precede the nouns they are attached to (it is obviously stretching the definition a bit to call these indeclinable determiners, but for the purposes of this post it’ll do).

The strong endings are taken by adjectives which come after these indeclinable determiners, and also by adjectives which come after no determiner at all. Note that there are a few adjectives like viel ‘much, many’, wenig ‘few’, einige ‘some’ and mehrere ‘several’ which have very determiner-like meanings, but are classified as adjectives on the basis of the endings that they exhibit (although many of these adjectives exhibit idiosyncrasies, e.g. viel and wenig take no ending in the singular when no determiner comes before them). Adjectives which come after other adjectives are inflected as if the preceding adjective was not present; what is important is the determiner (or lack thereof) which comes in front of the sequence of adjectives.

So, that’s the situation which we’re trying to understand here. Now, we already noted the high degree of syncretism in the weak adjective endings, which makes them of little use for the purpose of marking case, gender and number. The second important thing to note is that there is a one-to-one correspondance between the three kinds of determiners, as classified by the endings that they take, and the endings that adjectives that come after them take (weak, mixed and strong, respectively). For example, there are no determiners that decline like der but induce strong adjective endings, nor are there determiners are indeclinable but induce weak adjective endings. This suggests that there is some sort of causal link connecting the determiner endings and the adjective endings. And it’s not too difficult to see what this causal link might be. Think about how effectively each set of endings marks case, gender and number; let’s call this quantity marking capability. (This is not a standard term.) The weak adjective endings have a low marking capability. The strong adjective endings and both sets of determiner endings, on the other hand, exhibit relatively little syncretism and therefore have a high marking capability. The mixed adjective endings have a high marking capability in the nominative and accusative cases, but a low marking capability in the genitive and dative cases. And the indeclinable determiners, which take no endings, do not mark case, gender and number at all and therefore have zero marking capability. Let’s put all of these observations together in a table.

Determiner endings Adjective endings
der-like (high m.c.) weak (low m.c.)
ein-like (high m.c.) mixed (high m.c. in nom./acc., low m.c. in gen./dat.)
none (zero m.c.) strong (high m.c.)

You can see that there is a negative correlation between the marking capability of a set of determiner endings and the marking capability of the set of adjective endings that are associated with it. That is, if a determiner has a high marking capability, an adjective that comes after it will have a low marking capability, and vice versa.

So, we might explain the presence of two sets of adjective endings in German like this. Determiner-adjective sequences that come before nouns agree with nouns as units. The case, gender and number of the noun needs to be marked, but it can be marked either on the determiner or the adjective (or both); either way, it satisfies the requirement that the determiner-adjective sequence should mark the noun’s case, gender and number. When the determiner does the marking, there is no need for the adjective to do the marking as well—hence adjectives after der-like determiners take weak endings. When the determiner doesn’t do the marking, the adjective needs to do the marking instead—hence adjectives after indeclinable determiners take strong endings.

An extension of this approach also allows us to explain why the strong adjective endings in the genitive masculine singular and genitive neuter singular are -en (like the corresponding weak endings) and not -es (like the corresponding der-like determiner endings). In German, noun endings don’t mark case, in general, only number; case marking is left to the preceding determiner-adjective sequence. However, most masculine and neuter nouns take the ending -es in the genitive case. So, we can explain the use of the ending -en on both strong and weak adjectives before such nouns as due to the fact that the -es ending is already there on the noun.

This is the approach taken by many descriptions of German grammar; many learners of German as a second language find it illuminating, and it may well, to some extent, be how the system is understood by people learning German as their first language (it would be interesting to see if there have been any studies on this). However, it doesn’t explain everything. For example, the fact that adjectives after ein-like determiners take mixed endings is awkward to account for in this framework, since the ein-like determiners are no worse at marking case, gender and number in the nominative and accusative cases than in the genitive and dative cases (in fact, they’re probably better at it).

Can we take the analysis any deeper? Well, up until now, we’ve been examining the system from a synchronic rather than diachronic perspective, i.e. without regard for how it has changed over time. But it could be very much worth examining it from a diachronic perspective. The reason for this is that in general, languages tolerate a lot of contingent, more or less non-functional phenomena that are basically the result of historical accidents. For example, the plural of the word ox in English is oxen, rather than the expected *oxes. From a synchronic perspective, this is unexplainable; having this irregular plural doesn’t make the language more functional, all it does is make it necessary for learners to memorise an extra rule (“ox adds -en in the plural instead of -es”). But from a diachronic perspective, we can understand that oxen is a retention of the Old English plural oxan, with the ending -an, which was the regular plural ending of a class of nouns (the weak nouns) which was quite large in Old English. Over time, the alternative plural ending -as, used for another large class of nouns, was extended to more and more nouns. Historical linguists, drawing on studies that have been made of many different languages as they changed over time, know that changes like this (which are called analogical changes) tend to diffuse through the lexicon, affecting different words at different times—and this is what you would expect if you consider the nature of the processes, such as learner error, that drive this kind of change. So it is unsurprising that there is still one English word which is still holding out against the generalisation of the Old English -as plural ending. Of course, we can’t really answer questions like “why is it the word ox, in particular, which has retained this old ending?” It’s true that more common words are more likely to retain archaic traits, but that’s not a complete explanation. The diachronic perspective allows us to understand that we can’t expect this question to be answerable, since the diffusion process which causes words to acquire regularised plural endings in -es ~ -s is non-deterministic.

So, let’s have a look at how the system worked in the ancestor of the modern German language, which is known as Old High German, or OHG for short. It turns out that the system was quite similar in some respects. Just as in modern German, each OHG adjective took two different sets of endings depending on context; the endings from one set are known as the strong endings, and the endings from the other set are known as the weak endings, and indeed the modern German strong and weak adjective endings can be more or less straightforwardly derived from the OHG strong and weak adjective endings by regular sound change. The OHG forms of the strong and weak adjective endings are given in the table below.

OHG strong adjective endings
Singular Plural
Masculine Feminine Neuter Masculine Feminine Neuter
Nominative -ēr -iu -aʒ -e -o -iu
Accusative -an -a -aʒ -e -o -iu
Genitive -es -era -es -ero -ero -ero
Dative -emu -eru -emu -ēm -ēm -ēm
OHG weak adjective endings
Singular Plural
Masculine Feminine Neuter Masculine Feminine Neuter
Nominative -o -a -a -on -ūn -on
Accusative -on -ūn -a -on -ūn -on
Genitive -en -ūn -en -ōno -ōno -ōno
Dative -en -ūn -en -ōm -ōm -ōm

One important difference between the systems of OHG and modern German is immediately apparent from the tables above: the heavy syncretism of the weak adjective endings is more or less absent in OHG. Yes, they do exhibit some syncretism, but not significantly more than the strong adjective endings. In fact, if you compare the OHG weak adjective endings with the modern German strong adjective endings it is the latter which are more syncretic.

I can’t find much information on the circumstances which determined whether adjectives took strong or weak endings in OHG, although as far as I can tell, the key rules were the same as in modern German—adjectives took weak endings after the definite article dër and took strong endings when they did not come after a determiner. However, I do have access to more detailed information on the adjectives of another old Germanic language: Old English (OE). Although modern English has lost the distinction between the strong and weak endings entirely—indeed, it has lost all adjective endings except for the comparative and superlative endings, and even these are replaced by phrasal formation (more beautiful, most beautiful) for many adjectives—OE did inflect adjectives with two different sets of endings which are clearly cognate to the strong and weak adjective endings of OHG. The OE adjective endings are given in the two tables below.

OE strong adjective endings
Singular Plural
Masculine Feminine Neuter Masculine Feminine Neuter
Nominative -∅ -(u) -∅ -e -a -(u)
Accusative -ne -e -∅ -e -a -(u)
Genitive -es -re -es -ra -ra -ra
Dative -um -re -um -um -um -um
Instrumental -e -re -e -um -um -um
OE weak adjective endings
Singular Plural
Masculine Feminine Neuter Masculine Feminine Neuter
Nominative -a -e -e -an -an -an
Accusative -an -an -e -an -an -an
Genitive -an -an -an -ena -ena -ena
Dative -an -an -an -um -um -um
Instrumental -an -an -an -um -um -um

The bracketed endings in the table above only appear after stems which consist of a single light syllable, i.e. a syllable ending in a short vowel followed by a single consonant. This is due to a sound change which eliminated word-final short vowels after heavy syllables in pre-OE.

So, what determined the use of the strong or weak endings in OE? The following information is from Ringe & Taylor (2014).

First, note that unlike modern English or German, OE had no indefinite article. The word ān ‘one’, out of which the modern English indefinite article developed, could be used as an indefinite article, but it was optional; see, for example, the sentence below, in which æþelum brydguman ‘noble bridegroom’ appears without an article before it.

ond heo wæs þær beweddedo æþelum brydguman.
and there she was married to a noble bridegroom.

There was a definite article, although it was identical in form to the distal demonstrative (‘that’), se. (The modern English definite article comes from the nominative masculine singular form of the distal demonstrative3, while the modern English singular distal demonstrative that comes from the neuter form þæt.) The appropriate inflected form of se always appeared before definite nouns, unless definiteness was already inherent due to the noun being a proper noun or a pronoun, or already marked by a possessive. That said, in the earliest Old English poetry (e.g. Beowulf), we do see some definite nouns which do not have se before them; evidently the use of se as a definite article had not fully developed at the time these poems were written.

In general, weak endings are used after the demonstratives, þes ‘this’ and se ‘that’ (including the latter when it is used as a definite article), possessives and proper nouns (e.g. Carl Fǣtta ‘Charles the Fat’), while strong endings are used elsewhere. Strong endings are sometimes seen after non-pronominal possessives, but weak endings are much more common in this environment. The rule is thus quite similar to the German one, although there is the notable difference that in German adjectives takes mixed endings, not weak ones, after pronominal possessives, and they always take strong endings, never weak ones, after non-pronominal possessives.

Now, what do demonstratives, possessives and proper nouns have in common? Well, we just mentioned that possessed nouns and proper nouns never come after the definite article, because they are inherently definite. Definite nouns can be defined, roughly, as those which are associated with a particular referent, as opposed to indefinite nouns, which represent members of a certain class of things, with the identity of the members within the class not having any relevance. In English, possessives tend to be used only with definite nouns; when the noun is indefinite, constructions along the lines of a friend of mine are often used instead. One can use possessives with indefinite nouns too, but the option of using an alternative construction is there and, at least to me, the alternative construction usually sounds more natural. This is probably due to the fact that the most prototypical nouns that occur after possessives are body parts and kinship terms, which tend to be unique and hence definite (everybody has a single biological father, and hence nobody says a father of mine unless they are using the word father to mean something other than a biological father). I don’t know for sure whether the same could be said of the Old English possessives, but the fact that they behave like demonstratives and proper nouns—both of which have inherently definite referents—with respect to adjective endings does suggest so.

It seems, then, that the Old English rule can be briefly described as follows: adjectives before definite nouns take weak endings, adjective before indefinite nouns take strong endings. Now, it would be interesting to see what endings adjectives take before those nouns in Beowulf which are definite but do not have it marked by the demonstrative se. None of the sources I’ve read give any actual examples, but here is an apparent example from the Electronic Introduction to Old English:

gomela Scylfing hrēas heoroblāc.
the old Scylfing fell, mortally wounded (lit. “sword-pale”).

Here, the adjective gomela, which modifies the definite noun Scylfing (the Scylfings were the members of a particular Swedish royal family, and the phrase gomela Scylfing here is clearly intended to refer to a Scylfing called Ongenþēow mentioned earlier in the text; phrases like this were used in epic poems to avoid repetition of a name; hence why we know the noun is definite) has the weak nominative masculine singular ending -a, rather than its strong counterpart -∅. So it does seem that the use of the strong and weak endings is in accordance with the stated rule: weak endings appear before definite nouns. Of course, the proper thing to do here would be to do a full investigation of every definite noun that does not have se before it in Beowulf, but as this is just a blog post, which is getting long enough already, we’ll leave it here.

We can now see the original function of the weak endings. As there was no definite article in early Old English, the choice of the weak endings rather than the strong endings was the only thing that marked a noun as definite or indefinite. The use of the weak endings was not determined by morphological and syntactical factors, as in modern German, but by semantics: you added weak endings to indicate that a noun was definite, strong endings to indicate that a noun was indefinite.

It is likely that this system where the weak endings function as markers of definiteness operated in the same way in Proto-Germanic itself. Both Gothic and Old Norse (the two most divergent well-attested members of the Germanic family; the others are part of the West Germanic subgroup) had turned the Proto-Germanic distal demonstrative *sa ‘that’ into a definite article, and so had already progressed to the stage where the use of strong or weak endings no longer expressed a semantic distinction in its own right but was determined by the form of the determiner before the adjective. This situation must have developed from a system like the one we see in early Old English; that’s why we reconstruct such a system for Proto-Germanic.

Of course, the system doesn’t go back to Proto-Indo-European, because none of the other Indo-European languages have anything like it (we already talked about Latin, for example). So, one last important question remains. We now know what purpose the differentation between these two sets of endings served. But how did this differentiation come about?

Let’s have a look at the Proto-Germanic weak adjective endings, as reconstructed by Ringe (2006).

Proto-Germanic weak adjective endings
Singular Plural
Masculine Feminine Neuter Masculine Feminine Neuter
Nominative *-ō̄ ? ? *-aniz *-ōniz *-ōnō
Accusative *-anų *-ōnų ? *-anunz *-ōnunz *-ōnō
Genitive *-iniz *-ōniz *-iniz *-anǭ̄ *-ōnǭ̄ *-anǭ̄
Dative *-ini *-ōni *-ini *-ammaz *-ōmaz *-ammaz
Instrumental ? ? ? *-ammiz *-ōmiz *-ammiz

Although some of these endings are not reconstructible, of those that are, we can see a clear pattern: in general, each ending consists of a suffix of the form *-Vn, where the V is one of the vowels *a, or *i, followed by an ending which is the same for each gender. There are a couple of exceptions: the nominative masculine singular ending is *-ō̄, with no *n-containing suffix at the start, and the ending of the nominative and accusative neuter singular is *-ō, not *-iz as in the nominative masculine and feminine plural or *-unz as in the accusative masculine and feminine plural. Also, the *-n at the end of the suffix assimilates to *-m before the *-m- that begins the dative and instrumental plural endings, and the resulting cluster *mm is simplified to *m after the long vowel .

Compare the following endings, which are the endings of a particular nominal declension in Proto-Germanic, the consonant-stem declension:

Singular Plural
Nominative *-s, *-z, *-∅ *-iz (neuter *-ō)
Accusative *-ų *-unz (neuter *-ō)
Genitive *-iz *-ǭ̄
Dative *-i *-maz
Instrumental ? *-miz

These endings are more or less precisely those that appear after the *n-containing suffixes in the weak adjective endings. The only discrepancy is with the weak nominative masculine singular ending *-ō̄, but this can be traced back to pre-Proto-Indo-European *-ons, with the regular nominative singular ending *-s; a Proto-Indo-European (PIE) phonological rule reduced this ending to *-ō, which became *-ō̄ in Proto-Germanic. The classical example of this rule is Sanskrit śvā́ ‘dog’ (< PIE *ḱwṓ), the nominative singular corresponding to the vocative singular śván (< PIE *ḱwón; nouns in the vocative singular took no ending in PIE). Indeed, there are a couple of consonant stems ending in *-n- that have survived in PGmc (*ḱwṓ only survives in a suffixed form, *hundaz < PIE *ḱuntós), and these show the same loss of the -ns ending, with compensatory lengthening of the preceding -o, in the nominative singular, such as *sēmō̄ ‘seed’ (< PIE *séh₁mō, from pre-PIE *séh₁mons).

The final question, then, is where this *n-containing suffix came from. As far as I can tell, there is general agreement that it is related to a suffix attested in Greek and Latin which is used to “individualize” nouns. This suffix was a source of a few Roman cognomina: Catō, which has the genitive singular Catōnis, is derived from the adjective catus ‘shrewd’ (so Catō means ‘the Shrewd’), and Strabō, which has the genitive singular Strabōnis, is a Latinization of Greek Strabṓn, from strabós, an adjective that describes somebody with a squint (so Strabṓn means ‘the Squint-Eyed’). Jasanoff (2002) reconstructs the PIE form of the suffix as amphikinetic *-on- ~ *-n-. I haven’t tried to understand the details of his derivation, so I’m not sure how exactly how he explains the distribution of the three different vowels that appear in the *n-containing suffix.

The sources I can find are vague on the matter of how, exactly, the nickname-forming meaning in Greek and Latin is presumed definiteness-indicating meaning in Proto-Germanic. All I can do is give you my ill-informed speculation. Perhaps the Greek and Latin meaning is the more original one, and the indication of definiteness developed due to the inherent definiteness of proper nouns. But I’m not very confident about this. What we need here is an example of a language where a definite article, or some other way of indicating definiteness, is known to have developed from a similar source.

Anyway, we do now have a general idea of how the strong and weak adjective endings have developed, despite this respect (and others which I haven’t touched on, like the origin of the German mixed endings) in which the details aren’t clear. It might be helpful to give a brief summary.

In Proto-Indo-European, the distinction between weak and strong adjective endings did not exist. The weak endings were formed by adding regular endings onto a suffix of the form *-Vn. This suffix, which is used to form nicknames in Greek and Latin, somehow came to indicate definiteness in pre-Proto-Germanic. So, for example, adding it to *fullaz ‘full’ produced *fullō̄ ‘the full person or thing’. I am glossing it here as if it were a noun, but in Proto-Germanic (and also in modern German), nouns and adjectives could be freely interchanged; *fullaz when used as a noun would have meant ‘a full person or thing’, and *fullō̄ could be used as an adjective, so it could be placed in front of *gumō̄ ‘person’ giving *fullō̄ gumō̄ ‘the full person’. Due to sound changes, the separation between the suffix and the regular endings added to it became less clear in the individual German languages, so we generally speak of two sets of adjective endings, the strong endings (the original4 ones) and the weak endings (the endings with the remnants of the *n-containing suffix at the start). In Old High German, the distal demonstrative dër ‘that’ started to be used as a definite article, which made the original semantic function of the weak adjective endings redundant. Instead, they became simply variants of the strong endings which appeared in particular circumstances. Later sound changes in the German language caused many of the weak endings to become identical in form, so that in the end they only took two different forms, -e and -en. This enabled the synchronic analysis of the strong endings as “compensating” for the lack of marking capability in the weak endings (and, although I haven’t actually looked into it, I would imagine that this has something to do with the development of a new set of mixed endings).

By putting the development of the weak and strong adjective endings in a diachronic perspective, I’m not intending to say that the synchronic analysis of the German adjective ending system given above is worthless; in fact it seems likely that the synchronic analysis is valid in the sense that it may approximate how the system is understood by learners, and it may be the best explanation for the more recent developments that have occured in the system. But I do find that taking the diachronic perspective as well gives me a much more complete sense of understanding. If I took only the synchronic perspective there would still be a lot of questions that would seem like they needed answering, the most obvious being that of why German speakers bother having the weak adjective endings in the first place—wouldn’t it be much easier to just, say, not inflect the adjectives in the contexts where they take weak endings? And why not just make every adjective take the strong endings—the redundancy is not a problem, for if it were then why make adjectives agree with the nouns they are associated with at all? With the diachronic perspective, these questions no longer trouble me, because I can see that the weak adjective endings once served to indicate a particular meaning (that of definiteness), and now that this function has now been made redundant, they have only remained distinctive because their continued distinctiveness brings no significant costs, even if it brings no significant benefits either. And I think much the same could be said for other linguistic phenomena; a diachronic perspective is always useful, perhaps essential, for understanding. I think this is the main reason why my interests in linguistics tend more to the historical side.

  1. ^ The first/second declension is named thus because the Latin adjective declensions are more or less the same as the noun declensions, but there is the difference that each adjective takes endings associated with each of the three genders, while each noun only take endings associated with one gender (the gender of the noun). The nouns which take the first declension are mostly feminine, the nouns which take the second declension are mostly masculine or neuter, and the nouns which take the third declension can be of any of the three genders. (There are a couple more declensions, but these three account for the majority of nouns.) With adjectives, we find that some of them take the endings of the second declension in the masculine and neuter and the endings of the first declension in the feminine, while others take the endings of the third declension in the masculine, feminine and neuter. We therefore say that the first kind of adjectives take the first/second declension, while the second kind of adjectives take the third declension. None of this has any relevance to the point of the post, but I figured you might want to know.
  2. ^ Articles are sometimes omitted before singular nouns in certain kinds of sentences, e.g. in ich bin Schriftsteller ‘I am a writer’, but the article always appears, according to my Oxford German Dictionary, if an adjective comes before the adjective, so one says ich bin ein bekannter Schriftsteller ‘I am a famous writer’, never *ich bin bekannter Schriftsteller.
  3. ^ se had two different initial consonants, depending on which form it appeared in: s in the nominative masculine singular and nominative feminine singular, þ in every other form. In late Old English the initial consonant þ was generalised to every form, and the resulting nominative masculine singular form þe is the ancestor of modern English the (þ was the Old English way of writing th).
  4. ^ Actually, the Proto-Germanic strong adjective endings are considerably different from the original PIE adjective endings; they have largely (or even wholly) been replaced by “pronominal” endings, those of the distal demonstrative *se and other related words. The extension of pronominal endings to other classes of words is a pretty common development across the Indo-European languages, although it seems to have been more complete with respect to the Proto-Germanic strong adjective endings than in most of the others. It is unclear whether this has anything to do with the development of the strong and weak endings, or whether it is a completely independent development.

A brief history of English kinship terminology

In modern standard English, the following basic kinship terms exist:

father, mother, uncle, aunt, cousin, brother, sister, nephew, niece, husband, wife, son, daughter.

Phrases consisting of multiple words and terms which are regularly derived from more basic words via prefixes like grand- or great- or suffixes like -in-law are not included in this list. cousin is included here on the basis of its sense of ‘first cousin, i.e. uncle or aunt’s child’, not its sense of ‘relative who is not a direct ancestor or descendant’. Gender-neutral terms like parent, sibling, spouse and child are not included because, when the gender of the referent is known, it is always preferable to use a gender-specific term in English, so these terms are not as basic as the gender-specific terms.

The English kinship terminology system is a perfect example of an Eskimo kinship terminology system. Eskimo kinship terminology is the kind of terminology expected in a bilateral society, where no distinction is made between patrilineal and matrilineal ancestry and where the emphasis is on the nuclear family.

In Old English, the system was different. Here are the basic kinship terms of Old English (from the Bosworth-Toller Anglo-Saxon Dictionary):

fæder ‘father’, fædera ‘paternal uncle’, faþu ‘paternal aunt’, mōdor ‘mother’, ēam ‘uncle, esp. maternal’, mōdriġe ‘aunt, esp. maternal’, brōþor ‘brother, sweostor ‘sister’, nefa ‘nephew, grandson’, nift ‘niece, granddaughter’, swēor ‘father-in-law’, sweġer ‘mother-in-law’, tācor ‘husband’s brother’, sunu ‘son’, snoru ‘daughter-in-law’, dōhtor ‘daughter’, āþum ‘son-in-law, sister’s husband’.

Note that the precise meanings of the Old English kinship terms are difficult to identify, because the historical evidence is often incomplete, and also there was probably variation over time and space. So there might be some more obscure words, and additional senses to the words listed above, that have not been listed here. The above list, therefore, should be taken as a close but not exact approximation of the Old English kinship terminology system. With this caveat in mind, the following differences from modern standard English can be observed.

  • A distinction is made between paternal and maternal uncles and aunts. There were specific terms for paternal uncles and aunts, fædera and faþu respectively. The other two terms, ēam and mōdriġe, appear to have not referred exclusively to maternal uncles and aunts respectively but they were chiefly used in this sense.
  • The terms nefa and nift, chiefly meaning ‘nephew’ and ‘niece’ respectively, could also be used in the sense of ‘grandson’ or ‘granddaughter’, respectively. Note that unlike the terms for uncles and aunts, maternal and paternal nieces and nephews were not distinguished, although it was possible to use more specific derived terms like brōþordōhtor ‘brother’s daughter’.
  • It’s hard to find information on the Old English terminology for cousins; it seems that it isn’t well-attested, and people disagree about what distinctions where drawn. So I haven’t included any of it here. But according to Bosworth-Toller swēor ‘father-in-law’ could be used to refer to male cousins of some kind and mōdriġe ‘maternal aunt’ could be used to refer to female cousins of some kind. The use of swēor to mean ‘cousin’ is especially interesting because it may indicate that the Anglo-Saxons practised some kind of cousin marriage.
  • Like many languages, Old English lacked basic terms for ‘husband’ and ‘wife’; the words for ‘man’ and ‘woman’, wer or ceorl and wīf or cwēn respectively, were used instead).
  • Old English had basic terms for ‘father-in-law’ and ‘mother-in-law’: swēor and sweġer respectively. It also had basic terms for ‘son-in-law’ and ‘daughter-in-law’: āþum and snoru. However, āþum had an additional meaning of ‘sister’s husband’, and in this sense it translates modern standard English brother-in-law. But brother-in-law can also mean ‘husband’s brother’, and Old English had an entirely distinct word for this sense: tācor. As for ‘sister-in-law’, Old English does not appear to have had any basic terms for this, whether in the sense of ‘wife’s sister’ or ‘brother’s wife’.

The Old English kinship system does not fit neatly into any of Morgan’s classifications. It resembles the Eskimo kinship terminology of modern standard English in that paternal and maternal nephews and nieces are not distinguished; however, it does make a distinction between paternal and maternal uncles and aunts which is more typical of a Sudanese kinship terminology system. The Old English system might be seen as a system in a state of transition between a Sudanese system and an Eskimo system. The nonexistence of a basic term for ‘wife’s sister’ and the existence of a basic term for ‘husband’s brother’ might be taken as an indication that Old English society was patrilocal.

The Proto-Germanic kinship terminology system is of course even more difficult to know about, because the language is not attested in writing. However, based on the evidence of the older Germanic languages (Gothic, Old Norse, Old English, Old Frisian, Old Saxon, Old Dutch and Old High German), we can reconstruct an approximation of the system. The following list is based on information in Lehmann (2005-2007), A Grammar of Proto-Germanic and Ringe (2006), From Proto-Indo-European to Proto-Germanic.

*fadēr ‘father’ (c.f. Goth. fadar, ON faðir, OHG fater), *mōdēr ‘mother’ (c.f. Goth. mōdar, ON móðir, OHG muoter), *nefō̄ ‘nephew, grandson’ (c.f. ON nefe, OHG nevo), *niftiz ‘niece, granddaughter’ (c.f. ON nipt, OHG nift), *brōþēr ‘brother’ (c.f. Goth. brōþar, ON bróðir, OHG bruoder), *swestēr ‘sister’ (c.f. Goth. swistar, ON systir, OHG swester), *swehuraz ‘father-in-law’ (c.f. Old Swedish svēr, OHG swehur), *swegrū ‘mother-in-law’ (c.f. Goth. swaíhra, ON sværa, OHG swigar), *taikuraz ‘husband’s brother’ (c.f. OHG zeihhur), *sunuz ‘son’ (c.f. Goth. sunus, ON sunr, OHG sunu), *snuzō ‘daughter-in-law’ (c.f. OHG snura), *duhtēr ‘daughter’ (c.f. Goth. daúhtar, ON dóttir, OHG tohter), *aiþumaz ‘son-in-law, brother-in-law’ (c.f. OHG eidum).

Note that in Gothic, there were two other words for ‘father’ and ‘mother’ besides fadar and mōdar: atta and áiþei. The first of these has a PIE ancestor, *átta (c.f. Greek átta, Latin atta, both respectful terms of address for elderly men, and Hittite attas ‘father’). Ringe (2006) reconstructs *attō̄ for Proto-Germanic. However, áiþei is of unknown origin. The resemblance to *aiþaz ‘oath’ (which has a cognate in Old Irish ōeth, but no other Indo-European cognates, so it is probably a loanword from an unknown language that entered both Celtic and Germanic) is suggestive, but it could also be entirely unrelated. áiþei may also be related to *aiþumaz, which is also of unknown origin; it has no known cognates in any non-Germanic Indo-European languages, or indeed in any non-West Germanic language.

There were probably terms for uncles, aunts and cousins in Proto-Germanic as well, but they are difficult to reconstruct. On the basis of OE ēam and OHG ōheim we can reconstruct Proto-West Germanic *auhaimaz ‘maternal uncle’. This appears to be a contraction of a compound *awahaimaz formed from *awaz ‘uncle, grandfather’ (< Proto-Indo-European *h₂éwh₂os) + *haimaz ‘home’. But this is a strange compound, because compounds in the modern Germanic languages and in Proto-Indo-European are head-final (for example, elephant shrew refers to shrews that are like elephants, not elephants that are like shrews). An *awahaimaz is a kind of uncle, so this compound appears to be head-initial. I have no idea why this is the case. The choice of this compound to denote the maternal uncle is also interesting. If *awahaimaz is interpreted as ‘uncle who lives in the same home’ that suggests that the Proto-West Germanic speakers actually had a matrilocal society. In a patrilocal society, wives move into their husband’s homes after marriage, leaving their brothers behind, so people tend to live in extended families with their paternal uncles rather than their maternal uncles. This might seem strange, because it is pretty clear that later Germanic society and earlier Proto-Indo-European society was patrilocal. But there is, in fact, a theory that societies in the process of state formation tend to pass through a temporary matrilocal stage. For more on this, see my post on Tumblr about matrilocal societies.

There are other indications that Proto-Germanic preserved a reflex of *h₂éwh₂os (maybe *awaz?). Old Norse had the words afi ‘grandmother’ and amma ‘grandmother’; amma is probably a nursery word, but Lehmann says afi is a reflexes of *h₂éwh₂os (although I don’t know why the word has -f- rather than -v-). Apparently a dative singular form awōn ‘grandmother’ is attested from Gothic, too, which would correspond to nominative singular *awō. This might be the descendant of a feminine derivative, *awō (< PIE *h₂éwh₂ah₂, if it goes back as far as that), of *awaz in Proto-Germanic.

What about the other Old English words for uncles and aunts? Well, all of them lack cognates outside of West Germanic. fædera and mōdriġe are clearly derivatives of the words for ‘father’ and ‘mother’ respectively; they were probably originally adjectives meaning ‘paternal, i.e. of a father’ and ‘maternal, i.e. of a mother’ respectively. faþu also seems to be some kind of derivative of the word for ‘father’, although I don’t know what process would turn *fadēr into *faþō. Note the apparent Verner’s Law alternation!

Old English had a word mǣġ ‘relative’, which is not a kinship term has defined here. Its cognate in Old High German, māg, also means ‘relative’. However, in Old Norse mágr was a general term meaning ‘male relative by marriage, i.e. son-in-law, brother-in-law, father-in-law’, and in Gothic mēgs meant ‘son-in-law’ specifically. This word has no cognates in other Indo-European languages, and it is possible that was a kinship term with the ON or Goth. meaning in Proto-Germanic; then again ‘relative’ might just as well be the original meaning, especially if *aiþumaz is Proto-Germanic.

As for Proto-Indo-European, there is even more uncertainty than with Proto-Germanic, but the following kinship terms can be reconstructed.

*ph₂tḗr ‘father’ (c.f. Tocharian B pācer, Sanskrit pitā́, Old Armenian hayr, Greek patḗr, Latin pater, Old Irish athair), *máh₂tēr ‘mother’ (c.f. Tocharian B mācer, Sanskrit mātā́, Old Armenian mayr, Greek mḗtēr, Lithuanian mótė, Old Church Slavonic mati, Latin māter, Old Irish máthair), *h₂éwh₂os ‘grandfather’ (c.f. Hittite ḫūḫḫas, Old Armenian haw, Latin avus), *bráh₂tēr ‘brother’ (c.f. Sanskrit bhrātā́, Old Armenian ełbayr, Greek phrátēr, Lithuanian brólis, Old Church Slavonic bratrŭ, Latin frāter, Old Irish bráthair), *swésōr ‘sister’ (c.f. Tocharian B ṣer, Sanskrit śvasā́, Lithuanian sesuõ, Old Church Slavonic sestra, Latin soror, Old Irish siur), *swéḱuros ‘father-in-law’ (c.f. Sanskrit śvaśura, Greek hekurós, Albanian vjehërr, Old Church Slavonic svekrŭ ‘husband’s father’, Latin socer), *sweḱrúh₂ ‘mother-in-law’ (c.f. Sanskrit śvaśrū́s, Greek hekurā́, Old Church Slavonic svekry, Latin socrus), *dayhₐwḗr ‘husband’s brother’ (c.f. Sanskrit devā́, , Old Armenian taygr, Greek daḗr, Old Church Slavonic děverĭ, Latin lēvir), *yénh₂tēr ‘husband’s brother’s wife’ (c.f. Sanskrit yā́tṛ, Greek enátēr, Lithuanian jéntė, Old Church Slavonic jętry), *ǵh₂lōws ‘husband’s sister’ (c.f. Greek gálōs ‘sister-in-law’, Old Church Slavonic zŭlŭva ‘husband’s sister’, Latin glōs ‘husband’s sister’), *suHnús / *suHyús ‘son’ (c.f. Tocharian B soy, Sanskrit sūnú, Greek huiús, Lithuanian sūnùs, Old Church Slavonic synŭ), *népōts ‘grandson’ (c.f. Sanskrit nápāt, Greek anepsiós ‘cousin’, Albanian nip ‘grandson, nephew’, Old Church Slavonic netijĭ ‘nephew’, Latin nepōs ‘grandson, nephew’, Old Irish nïa ‘sororal nephew’), *snusós ‘daughter-in-law’ (c.f. Sanskrit snuṣā́, Old Armenian nu, Greek nuós, Latin nurus), *dʰugh₂tḗr ‘daughter’ (c.f. Tocharian B tkācer, Sanskrit duhitā́, Old Armenian dustr, Greek thugátēr, Lithuanian duktė̃, Old Church Slavonic dŭšti)

There were probably feminine counterparts to *h₂éwh₂os and *népōts in Proto-Indo-European, but they were formed as derivatives of the masculine terms. There are numerous indications that the society of the Proto-Indo-European speakers was patrilocal: swéḱuros seems to have referred to a husband’s father only, not a wife’s father, there is a basic term for ‘husband’s brother’s wife’ but not ‘husband’s sister’s wife’, and it is uncertain whether there are reconstructable basic PIE terms for ‘wife’s brother’ or ‘wife’s sister’.

It looks to me like the evidence of kinship terminology suggests that English-speakers and their linguistic ancestors have been patrilocal for most of their history. That said, as mentioned above, Proto-Germanic *awahaimaz suggests that there might have been a short matrilocal period around the Proto-Germanic period. This is far from conclusive evidence on its own, but there are also clues that the Germanic peoples might have been to some degree matrilocal (or avunculocal) from Tacitus’s Germania, and if the Harris-Divale theory of matrilocality being related to external warfare during state formation is correct, this would be a prediction of that theory.

Some facts about gender

One of the most interesting phenomena found in languages is gender. In its linguistic sense, gender refers to the phenomenon where nouns are divided into a number of different classes which can be distinguished due to the fact that words associated with nouns, like pronouns, determiners, adjectives and verbs, often appear in different forms depending on the gender of the nouns they are associated with (this is called gender agreement). For example, in German the word for ‘the’ is der when it is attached to masculine nouns like Mann ‘man’, die when it is attached to feminine nouns like Frau ‘women’ and das when it is attached to neuter nouns like Kind ‘child’. The main reason gender is such an interesting phenomenon is probably that is not at all obvious why it exists. Language is generally thought of as a means of communication, but it is hard to see how gender systems aid communication. Even if there might be some benefits, any explanation has to account for the high prevalence of gender systems in languages worldwide: in the WALS‘s sample 112 out of 257 languages, about 44%, make some kind of distinction between two or more genders.

Something people aren’t always aware of is that English has gender as well. In fact, it has three genders, like German: masculine, feminine and neuter. Admittedly, there are two differences between the gender system of English and the gender systems of languages like French and German which make gender a less prominent phenomenon in English.

Firstly, in English, gender agreement only occurs with pronouns. The words he, she and it are used to refer to males, females and non-gendered things, respectively, and using the wrong pronoun for a given referent is considered grammatically incorrect.1 On the other hand, in French and German gender agreement also occurs with determiners and adjectives, and in written French, and in both spoken and written Russian gender agreement also occurs with verbs. Languages like English where gender agreement only occurs with pronouns are said to have pronominal gender systems. But promoninal gender systems are gender systems nonetheless. Remember above I said that only 44% of the languages in the WALS’s sample distinguish two or more genders: well, English and other languages are counted among that 44%, so a majority—56%—of the languages in the sample show no gender agreement, not even in pronouns. In fact, pronominal gender systems are quite rare, and it is likely that most of them are the result of not-quite-complete loss of an original, more extensive gender system. This is certainly the case for English. I think it’s quite possible that in the future, English will lose the last vestiges of its gender system as people switch to using they to refer to people without regard to gender in all circumstances, as they already tend to do when the gender of a person is unknown.

Secondly, English gender assignment corresponds almost exactly to the meaning of the referent. Perhaps the only exception is that ships are often referred to as she, but even this is optional. On the other hand, in French and German there are many things which do not have a gender but are classified as masculine or feminine. It’s not the case that speakers of these languages define gender in a different way from English speakers: French people do not actually think that curtains are male and tables are female, even though they say un rideau, not *une rideau and une table, not *un table. And in German, Gardine ‘curtain’ is feminine and Tisch ‘table’ is masculine, so it seems unlikely that the choice of genders for these objects is based on any inherent association of them with masculinity or femininity given that two neighbouring peoples sharing similar cultures have made the assignments in two completely different ways. The French and German masculine and feminine genders just contain a lot of other things besides males and females. In fact, in German, there is an example which goes the other way. Mädchen ‘girl’ should be feminine given the meaning2, but it is actually neuter: Germans say das Mädchen, rather than *die Mädchen.3

There is, however, an important difference between the assignment of Mädchen to the neuter gender and the assignment of Tisch and Gardine to the masculine and feminine genders respectively. The reason Mädchen is neuter is that it it is formed from the word Magd ‘maiden’ by adding the dimunitive suffix -chen, and there is a rule in German that says that every word formed by adding the suffix -chen is neuter. Many German suffixes are associated with a particular gender; for example, nouns in -heit, -keit and -schaft are always feminine, and nouns in -lein (which is another dimunitive suffix) are always neuter. The associations of these suffixes with particular genders constitute a rule which overrides the rule that every word that refers to males is masculine and every word that refers to females is feminine. So, in German gender assignment is determined by (at least) two rules, which are applied in sequence.

  1. If the noun is formed with the suffixes -heit, -keit or -schaft, assign it to the feminine gender, and if the noun is formed with the suffixes -chen or -lein, assign it to the neuter gender. (Note: there are other suffixes which should be mentioned in this rule, as well, but I’m not intending to precisely describe German gender assignment here, just to show you the general outline of how the system works.)
  2. If the noun refers to males, assign it to the masculine gender. If the noun refers to females, assign it to the feminine gender.

Note that the first rule is formal in nature (it refers to how the words are formed) while the second rule is semantic in nature (it refers to what the words mean). The existence of formal rules is responsible for another way in which English gender assignment is different from French and German assignment. In English, gender is a property of referents, not of the nouns themselves. But in French and German, gender is more a property of the nouns themselves. In these languages, it is possible for the same thing to be referred to by two different nouns of different genders. For example, in French, the word vélo is masculine and the word bicyclette is feminine; this is because bicyclette ends in the feminine dimunitive suffix -ette.

These rules are not sufficient to assign every German noun to a gender. Of course, this is not meant to be a complete list. But there is an interesting question here: can a list of rules based on formal and semantic factors account for the gender assignment of every noun in German? Or are there some words whose gender assignment is simply arbitrary?

If you have any knowledge of German, you might find it hard to believe that gender assignment is not mostly arbitrary. People who know French would probably also expect gender assignment in that language to also be mostly arbitrary. However, quite a few studies have been carried out which have shown that gender in French is mostly predictable via phonological rules: that is, rules that take into account the sound of the word. For example Tucker, Lambert & Rigault (1977) found that 94% of French nouns ending in the sound /ʒ/ (such as ménage ‘housekeeping’) are masculine. There are exceptions: orange ‘orange’ is feminine. But by using rules like this, Tucker, Lambert & Rigault were able to correctly determine the genders of 85% of the nouns in the Petit Larousse, a famous French dictionary. Since they did not take into account semantic (i.e. relating the meanings of words) and morphological (i.e. relating to the composition of words from prefixes, suffixes, etc.) factors, and the phonological rules they found could probably be made more accurate, it is quite likely that the vast majority of French nouns are predictably gendered. Given this surprising result, it is possible that a lot more of German assignment is predictable than you might think. Köpcke & Zubin (1984) were able to find a large amount of regularity using similar techniques, although the rules appear to be more complex than the rules in French. Of course, the more complex the assignment rules are, the less useful they are for prediction because it is difficult for learners to remember them all and apply them quickly. There is no bright line between gender being hard to predict and gender being completely unpredictable, since if you just have one rule for every word in the language saying “this word is masculine / feminine / neuter”, then that is still a set of rules. The conclusion I would draw from results like those of Tucker, Lambert & Rigault is that French and German gender is more predictable than you might think, even though it is often not fully predictable in practice.

Few gender systems have been studied as much as the French and German gender systems have, so it is possible that we might find languages that have significantly more unpredictable gender assignment rules. But it would be surprising, since predictable assignment rules are a lot more convenient for learners. I think it’s more likely that in all languages that distinguish different genders, gender assignment is to a large degree predictable.

Another interesting example of a language with apparently unpredictable gender assignment is Ojibwa. In Ojibwa there are two genders which are called the animate and inanimate genders. These genders have nothing to do with sex (the word gender comes from the French word genre, which just means type; it can refer to any kind of distinction between nouns that is reflected in agreement). Nouns that denote people, animals, trees or supernatural beings are always animate. Most other nouns are inanimate. But there is a fairly large group of nouns that seem like they should be in the inanimate gender, but are actually animate. These include ekoːn ‘snow’, enank ‘star’, esseːmaː ‘tobacco’, mentaːmin ‘maize’, meskomin ‘raspberry’ and ekkikk ‘kettle’ (Bloomfield 1957). Now, some of these might be explainable as resulting from differences in which things are considered to possess a gender. For example, it is very common, cross-linguistically and cross-culturally, for celestial bodies such as stars to be identified with supernatural beings, who have a gender. And given that trees are considered animate in Ojibwa, it’s possible that other plants like tobacco and maize might be considered animate as well. But other examples like ekoːn ‘snow’ being animate are harder to explain. There is no generally-accepted explanation for the composition of the Ojibwa animate gender, but Black-Rogers (1982) has an interesting one. According to Black-Rogers, the Ojibwa lack a clear distinction between natural and supernatural abilities. They believe that even fairly mundane activities like beadwork are only possible because of powers that have been granted to humans via supernatural means. Inanimate objects, in particular, may be sources of power. Different speakers may disagree as to which objects have power, and power may be considered to come from different sources at different times. Black-Rogers proposed that when an object is considered to be a source of power speakers start assigning it to the animate gender. She was able to explain many of the problematic animate nouns by this means. For those she wasn’t able to explain, she suggests that objects assigned to the animate gender tend to stay there, so there may be animate nouns which refer to objects that were formerly considered to be sources of power, but no longer are today. So Ojibwa gender assignment is to some extent arbitrary from a synchronic perspective, but in diachronic perspective it can be completely explained by semantic factors. Black-Rogers’s explanation may or may not be true, but I brought up the example to show you that highly unpredictable gender assignments can also be influenced mainly by semantic factors, rather than by formal factors as in the case of French and German.

Another interesting question about gender is whether there are any languages where gender assignment is determined entirely by formal factors, so that semantic factors are irrelevant. Now, it’s true that in many languages formal rules can almost entirely predict gender assignment. In Hausa, for example, there is a very simple rule: nouns ending in -aa are feminine, and all other nouns are masculine. There are some exceptions to the rule, but they are few in number. However, semantic factors are not irrelevant. If they were, then we would not expect nouns referring to males to be masculine and nouns referring to females to be feminine, because there is no reason why nouns referring to females should end in -aa but other nouns should not. In fact, the vast majority of nouns referring to females end in -aa and are feminine. Historically, the correlation between the feminine gender and the -aa suffix did not exist, or was less strong. What happened was that a suffix -nyàa was used to form nouns denoting females, and this resulted in -aa being associated with nouns denoting females, so that all such nouns ended up having the suffix -aa added to them.

Hausa gender, then, is not determined only by formal factors, and in fact it seems that there are no languages where gender is determined only by formal factors. In general, gender distinctions seem to always be fundamentally based on semantic rules of the form “all nouns with meanings of type A are assigned to gender X” (so the gender X contains all nouns of type A, but not necessarily only nouns of type A). These rules give each gender an initial set of nouns called its “semantic core”. Then semantic associations and formal rules are sufficient to assign the vast majority of remaining nouns to one of the genders, and they may shift nouns within the semantic core of one gender to a different gender as well.

So, I’ve talked a bit about about gender assignment and the interaction between formal and semantic factors here, but there are lots more interesting things to talk about with respect to gender in languages, such as: what kind of distinctions tend to be drawn? Male vs. female, animate vs. inanimate are very common—any others? What can borrowings tell us about gender assignment? What can we say about nouns which appear to have characteristics of multiple genders (like German Mädchen)? How do gender systems develop and change over time? How are gender systems acquired by language learners? If you’re interested, I recommend Gender (1991) by Greville Corbett. This post is based on the first few chapters of that book.


  1. ^ There is a known phenomenon where English speakers sometimes refer to things that would normally be referred to by it by he or she instead; for example a teenage boy told a surfer, referring to a wave: “Catch her at her height!” (Corbett 1991). But this occurs only in particular circumstances; it is clearly the usual pronoun used to refer to these things.
  2. ^ It is common, cross-linguistically and cross-culturally, for children to be non-gendered, and indeed Kind ‘child’ is neuter; however, Junge ‘boy’, and its older synonym Knabe are both masculine, so we would expect Mädchen to be feminine in parallel.
  3. ^ However, in colloquial German pronouns often agree with Mädchen as if it was feminine: Kennst du das Mädchen? Nein, ich kenne sie nicht, not … kenne es nicht.


Black-Rogers, Mary B. 1982. Algonquian Gender Revisited: Animate Nouns and Ojibwa ‘Power’ – an Impasse? Papers in Linguistics 15. 59-76.

Bloomfield, Leonard. 1957. Eastern Ojibwa: Grammatical Sketch, Texts and Word List. Ann Arbor: University of Michigan Press.

Corbett, G. G. 1991. Gender. Cambridge: Cambridge University Press.

Köpcke, K. M. & Zubin, D. 1984. Sechs Prinzipien für die Genuszuweisung im Deutschen: Ein Beitrag zur natürlichen Klassifikation. Linguistische Berichte 93: 26-50.

Tucker, G. R., Lambert, W. E., & Rigault A. A. 1977. The French speaker’s skill with grammatical gender: An example of rule-governed behavior. The Hague: Mouton de Gruyter.

The phonetic motivation for Grimm’s Law

…is not as clear as I had thought.

According to the standard reconstruction of Proto-Indo-European (PIE), the language had three series of stops. One of the series is thought to have consisted of voiceless unaspirated stops: *p, *t, *ḱ, *k, and *kʷ. Another is thought to have consisted of voiced unaspirated stops: *b, *d, *ǵ, *g and *gʷ. And the other is thought to have consisted of voiced aspirated stops: *bʰ, *dʰ, *ǵʰ, *gʰ and *gʷʰ. These series were preserved in this form in Sanskrit, although Sanskrit also innovated a fourth series of voiceless aspirated stops out of clusters consisting of voiceless stops followed by laryngeals. In Proto-Germanic, however, the situation is different. The PIE voiceless unaspirated stops have become voiceless fricatives; c.f. Proto-Germanic *þū (> English thou) and Sanskrit tvám ‘you (singular)’. The PIE voiced unaspirated stops have become voiceless unaspirated stops; c.f. Proto-Germanic *twō (> English two) and Sanskrit dvā́ ‘two’. And the PIE voiced aspirated stops have become voiced unaspirated stops; c.f. Proto-Germanic *meduz (> English mead) and Sanskrit mádhu ‘honey’.

I had always assumed that the change went something like this. First, the voiceless unaspirated stops fricativised, retaining their lack of voice and aspiration and becoming voiceless fricatives. Changes of stops into fricatives are common and unremarkable; phonologists disagree on whether this is due to a natural tendency towards lenition (weakening) or due to assimilation to neighbouring phonemes which are more sonorous, but there is no dispute that such a change can be phonetically motivated. The change can be written formally in terms of distinctive features as follows.

[-continuant, -voice] > [+continuant]

Second, the voiced unaspirated stops devoiced, retaining their lack of frication and aspiration and becoming voiceless unaspirated stops. This change would be unusual if it occured on its own. However, the previous change had left the language with no voiceless unaspirated stops, only voiced unaspirated stops and voiced aspirated stops. The [±voice] feature which had been used to distinguish the three original series was now redundant. For obstruents the unmarked value of this feature is [-voice] (voicelessness); that is, obstruents tend to be voiceless unless something forces them to be voiced. Therefore, it was natural for the voiced unaspirated stops to be devoiced. The change can be written formally in terms of distinctive features as follows.

[-continuant, +voice, -spread glottis] > [-voice]

Third, the voiced aspirated stops deaspirated, retaining their voicing and lack of frication and becoming voiced aspirated stops. This change was increased in likelihood due to the fact that the previous change left the language with two series of stops, one of which was voiceless unaspirated and one of which was voiced aspirated; the two features [±voice] and [±spread glottis] were therefore redundant against each other. [-spread glottis] is the unmarked value of the [±spread glottis] feature on stops, so it was natural to resolve this by deaspirating the voiced aspirated stops (although devoicing the voiced aspirated stops would have worked just as well). The change can be written formally in terms of distinctive features as follows.

[-continuant, +spread glottis] > [-spread glottis]

But there are two questions I have about this account.

  1. Why did the second change involve the voiced unaspirated stops devoicing, rather than the voiced aspirated ones? The redundancy of the [±voice] feature could have been resolved either way. In fact, why didn’t both kinds of stop devoice? Since [-voice] is unmarked for obstruents there is nothing stopping this from happening.
  2. Why did the third change involve the voiced aspirated stops deaspirating rather than devoicing? Since the [±voice] and [±spread glottis] features were redundant against each other devoicing would have worked just as well as a means of resolving the redundancy.

Now, sound change is not a deterministic process, so perhaps the answers to these questions are just that out of all of the different ways the redundancies in question could be resolved, these were the ways that were chosen, more or less at random. I am satisfied with this as the answer to question 2. In fact, with respect to question 2 it seems like deaspiration would be a more likely occurence than devoicing because it is much more common for languages to distinguish stops using the [±voice] feature than it is for them to distinguish stops using the [±spread glottis] feature; contrasts of voice are therefore probably favoured over contrasts of aspiration (although this is only a tendency, and there are plenty of languages like Mandarin Chinese where [±spread glottis] is distinctive but [±voice] is not).

But I am less satisfied with this as an answer to question 1. As I mentioned above, the redundancy of the [±voice] feature could have been solved in three different ways:

  1. devoicing of the voiced unaspirated stops, resulting in a contrast between voiceless unaspirated stops and voiced aspirated stops.
  2. devoicing of the voiced aspirated stops, resulting in a contrast between voiced unaspirated stops and voiceless aspirated stops.
  3. devoicing of both kinds of stop, resulting in a contrast between voiceless unaspirated stops and voiceless aspirated stops.

There are languages with a contrast between voiced unaspirated stops and voiceless aspirated stops, as would result from option 2. English is such a language. There are also languages with a contrast between voiceless unaspirated stops and voiceless aspirated stops, as would result from option 3. Mandarin Chinese is such a language. But I know of no language which has a contrast between voiceless unaspirated stops and voiced aspirated stops, as would result from option 1. Yet option 1 seems to have been the option that was taken. This is odd.

I think there are phonetic reasons why we would expect options 2 or 3 to be favoured over option 1. If you examine the articulatory mechanisms which are used to produce voiced aspirated stops, you can see them as half-voiced stops, closer to voiceless stops than voiced unaspirated stops (but still voiced). If you think about voiced aspirated stops in this way, option 1 is weird, because it involves change of the voiced unaspirated (i.e. fully voiced) stops directly into voiceless unaspirated stops without passing through the intermediate stage where they would be voiced aspirated (i.e. half-voiced) and end up merging with the voiced aspirated stops. If the characterisation of voiced aspirated stops as half-voiced already makes sense to you, you can skip the next few paragraphs, because I’m now going to try and explain why this is an accurate characterisation.

The first thing that I want to explain is what voiced aspirated stops are. In terms of distinctive features, they are parallel to voiceless aspirated stops. Voiced aspirated stops are [+voice] and [+spread glottis], voiceless aspirated stops are [-voice] and [+spread glottis]. But the meaning of [+spread glottis] is different in the two cases. As a feature of voiceless stops, [+spread glottis] corresponds to increased duration of the period during which the vocal folds are prevented from vibrating (normally by keeping the vocal folds apart from each other, hence the name of the feature, although reducing the airflow is also an option). The between the release of a stop and the beginning of vocal fold vibration in order to voice the following voiced phoneme is called the voice onset time (VOT). For voiceless unaspirated stops, the VOT is close to 0, while for voiceless aspirated stops the VOT is larger, so that there is an audible period after the stop has been released where air flows through the glottis but the vocal folds do not vibrate. This results in a sound being produced during this period which is in fact exactly [h], the voiceless glottal continuant (although speakers of languages which have aspirated stops don’t usually perceive the [h], instead perceiving it as part of the preceding stop).

During the production of voiced stops, the vocal folds are already vibrating (that’s what it means for a stop to be voiced). So it is impossible for voiced stops to be aspirated if aspiration is defined as having a positive VOT1. Instead, [+spread glottis] as a feature of voiced stops corresponds to the vocal folds being held further apart than is normal for voiced stops, roughly speaking. The vocal folds are still close enough that they vibrate during the production of voiced aspirated stops, so such stops are not completely voiceless, but they are closer to voiceless than voiced unaspirated stops. The kind of voice that accompanies voiced aspirated stops is called breathy voice, as opposed to the modal voice that accompanies voiced unaspirated stops. It might help to look at the following diagram, which illustrates the relationship between the degree of closure of the glottis and different kinds of voicing. The diagram is adapted from Gordon & Ladefoged (2001).

Voiceless sounds have the least glottal closure. The glottal stop has the most glottal closure (complete closure). Modally-voiced sounds have a degree of glottal closure midway between these two extremes. Breathy-voiced sounds have a degree of glottal closure between that of voiceless sounds and voiced sounds. Creaky-voiced sounds have a degree of glottal closure between that of voiced sounds and the glottal stop.

(I should note that talking about the degree of closure of the glottis as if this was a scalar variable is an oversimplification. When the vocal folds vibrate, what happens is that the glottis alternates between a state where it is more or less fully open (as when a voiceless sound is being produced) and a state where it is more or less fully closed (as when a glottal stop is being produced). Closure occurs due to tension from the laryngeal muscles and opening occurs due to pressure from the flow of air through the trachea; closure results in buildup of air below the glottis, resulting in increased pressure, while opening allows air to flow through a greater area, resulting in decreased pressure, and this is why the alternation occurs. For a given rate of flow of air, there is a maximal tension above which opening cannot occur and a minimal tension below which closure cannot occur, and in between these two extremes there is an optimal tension which results in maximal vibration; this tension is approached during the production of modally-voiced sounds. If the tension is below the optimal tension but above the minimal tension, the result is a breathy-voiced sound. If the tension is above the optimal tension but below the maximal tension, the result is a creaky-voiced sound. Alternatively, creaky-voiced sounds can be produced by having the glottis completely closed at one end, with modal voice at the other end, and breathy-voiced sounds can be produced by having the glottis open so that the vocal folds do not vibrate at one end, with modal voice at the other end. But regardless of how these sounds are produced, they sound the same, so the distinction is not important. Either way, it is still accurate to say that breathy-voiced sounds are in a position between voiceless sounds and modally-voiced sounds.)

It would be helpful to see how voiced aspirated stops behave with respect to sound change in attested languages. Unfortunately, voiced aspirated stops are rare. which limits the number of available examples. As far as I know, voiced aspirated stops are mainly found in the Indo-Aryan languages of South Asia and the Nguni languages of South Africa. In the Indo-Aryan languages the voiced aspirated stops have been inherited from PIE, or at least Vedic Sanskrit (depending on what you believe about the nature of the PIE stops), and most of them seem to have preserved them unchanged. Sinhala and Kashmiri have no voiced aspirated stops, but I don’t know and can’t find any information on what happened to them in these languages. So it seems that the voiced aspirated stops have been stable in these languages. That suggests the rarity of voiced aspirated stops is probably more due to the infrequency of sound changes that would make them phonemic rather than inherent instability. However, the mutual influence of these languages upon each other within the South Asian linguistic area might have helped preserve the voiced aspirated stops; the fact that the two most peripheral Indo-Aryan languages do not have them is perhaps suggestive that this has been the case. What about the Nguni languages? These are a tight-knit group, probably having a common origin within the last millennium, and their closest relatives such as Tswana have no voiced aspirated stops. So their voiced aspirated stops are of more recent vintage. Interestingly, Traill, Khumalo & Fridjhon (1987) have found that the Zulu voiced aspirated stops are actually voiceless, with the breathy voice occuring after the release on the following vowel. This seems like it could be the first step on a change of voiced aspirated stops into voiceless aspirated stops. But I don’t think any of this evidence is of much use in making the case that Grimm’s Law is weird. My case primarily rests on the idea that voiced aspirated stops are intermediate between voiceless and modally-voiced stops on the basis of how they are produced.

If the changes as described above are odd, maybe we should consider the possibility that the changes described by Grimm’s Law were of a different nature.

Perhaps a minor amendment can solve the problem. It is universally agreed that the Proto-Germanic voiced stops had voiced fricative allophones. It is not totally clear which environments the stops occured in and which environments the fricatives occured in, but they were all definitely stops after nasals and when geminate and fricatives after vowels and diphthongs. There are three different ways this situation might have come to be.

  1. The PIE voiced aspirated stops might have turned into voiced unaspirated stops first and then acquired fricative allophones in certain environments.
  2. The PIE voiced aspirated stops might have turned into voiced unaspirated fricatives first and then acquired stop allophones in certain environments.
  3. The PIE voiced aspirated stops might have turned into voiced unaspirated fricatives in certain environments and voiced unaspirated stops in others.

If we suppose that number 2 is the accurate description of what happened, then it is possible that the fricativisation of the PIE voiced aspirated stops occured before the devoicing of the PIE voiced unaspirated stops. This devoicing would then be perfectly natural because the PIE voiced unaspirated stops would be the only stops remaining in the language, so the marked [+voice] feature would be dropped from them. The voiced aspirated stops would probably have become voiced aspirated fricatives (i.e. breathy-voiced fricatives) initially and then these fricatives would have become modally-voiced since there would be no need for them to contrast with modally-voiced fricatives. Is it plausible that the voiceless unaspirated and voiced aspirated stops would have devoiced, but not the voiced unaspirated stops? What do these two kinds of stop have in common that the third stop lacks? If we think of voiced aspirated stops as half-voiced stops, we can describe the change as affecting all of the stops which were not fully voiced. The change is especially plausible, however, if we suppose that the PIE voiceless unaspirated stops had become aspirated before the changes described by Grimm’s Law took place. In that case, the change would affect the aspirated stops and not affect the unaspirated stops. Fricativisation of aspirated stops but not unaspirated stops is a very well-attested sound change; it happened in Greek, for example. The sequence of changes would be as follows:

[-continuant, -voice] > [+spread glottis]

[-continuant, +spread glottis] > [+continuant]

[-continuant, +voice] > [-voice]

[+continuant, +voice] > [-continuant]

(The last change would have occurred only in some environments; there are also conditioned exceptions to some of the other changes.)

Is there any other reason to think the PIE voiceless unaspirated stops might have become aspirated in Proto-Germanic before fricativising? Well, the reflexes of the Proto-Germanic voiceless stops are aspirated in the North Germanic languages and English, and have become affricates in some positions in German which suggests that they were originally aspirated; the lack of aspiration in Dutch can probably be attributed to French influence. That suggests the Proto-Germanic voiceless stops were already aspirated. Of course, these voiceless stops are the reflexes of the PIE voiced unaspirated stops, not the PIE voiceless unaspirated stops. But perhaps the rule that aspirated voiceless stops was persistent in Proto-Germanic, so that it applied to both the PIE voiceless unaspirated stops before they fricativised and the PIE voiced unaspirated stops after they were devoiced. The rule seems to have persisted into German, because German went through its own kind of replay of Grimm’s Law in which the Proto-Germanic voiceless stops became affricates or fricatives and the Proto-Germanic voiced stops were devoiced. This second consonant shift was never fully completed in most German dialects; in Standard German, for example, Proto-Germanic *b and *g were not devoiced in word-initial position. However, *d was devoiced (c.f. English daughter, German Tochter) and modern Standard German /t/ is aspirated, so, for example, Tochter is pronounced [ˈtʰɔxtɐ].

I think this is a satisfactory solution to the problem. The idea that the PIE voiced aspirated stops became fricatives first is not a new one, in fact it is probably the favoured scenario, but I have never seen it justified in this way, and Ringe (2006) suggests that the voiced aspirates changed into both stops and fricatives depending on the environment (number 3 above), which is incompatible with the scenario I have proposed here.

Finally, I think I should mention that all of this reasoning has been done on the assumption that PIE had voiceless unaspirated, voiced unaspirated and voiced aspirated stops. If you subscribe to an alternative hypothesis about the nature of the PIE stops, such as the glottalic theory, Grimm’s Law might have to be explained in a completely different way. But despite it not being as easy as it might appear at first glance, it does seem that the standard hypothesis is capable of explaining Grimm’s Law.

Whether it can explain Verner’s Law is another matter. I have always thought it a little odd that the voiceless fricatives were voiced after unaccented syllables but not after accented syllables. It is not obvious how accent and voice can affect each other. But I’ll discuss this, perhaps, in another post.


Gordon, M., & Ladefoged, P. (2001). Phonation types: a cross-linguistic overview. Journal of Phonetics, 29(4), 383-406.

Ringe, D. (2006). From Proto-Indo-European to Proto-Germanic: A Linguistic History of English: Volume I. Oxford University Press.

Traill, A., Khumalo, J. S., & Fridjhon, P. (1987). Depressing facts about Zulu. African Studies, 46(2), 255-274.