Rules for German

Schlossplatz [‘castle plaza’] in Stuttgart, Germany. 📷: by author (2023).

[ Glossary ]

INTRODUCTION

In this post, we want to provide you with some tools to help you master some of the trickier points of German grammar. These include: (1) patterns in the 16 forms for the definite article ‘the’; (2) some remarks on the two major classes of German verbs, with reference to those of English; and (3) a fool-proof system generating correct word orders, and avoiding incorrect ones.

Los geht’s! We hope you’ll find this helpful, whether you’re learning this for the first time, or wanting to review the fundamentals. I want to thank LanGo’s 2024 class of German students, especially Wesley!

1 The ‘The’

As you may know, German and English are fairly closely related, as languages go: 15 to 20-ish centuries ago, our linguistic ancestors belonged to a single language community (see our blog posts “Natural Sound Change” and “Language Change: How Latin Became Spanish” for more on how and why language changes). This is the reason we share many grammatical rules and a large set of our vocabularies. This inherited word-stock shows evidence of (largely) regular sound changes: among these is that English <th> (pronounced [θ] as in “thin,” or [ð] as in “this”) regularly corresponds to German <d>. (Here the German development is an innovation, and the English fricatives are retentions from the proto-language.) German for ‘thin’ is dünn, and the stem for ‘this’ is dies-. We see this correspondence also in our words for the definite article, “the” and der/die/das.

Modern German has preserved, with minor changes, the ancestral inflection system, distinguishing 2 numbers (singular and plural), 3 genders (masculine, feminine, and neuter), and 4 cases[1] (see Table 1; German and English show traces of an old 5th case, the instrumental, as well). Table 1 presents all 16 forms a German learner needs to master as soon as possible. While the order of cases is traditional, I prefer to show masculine and neuter side by side, to emphasize the overlap of some of their forms.

Your mission, dear learner, is to memorize this table: be able to reproduce all the forms without error. A few points to help you with this …

1. Vowel length

As in English, short vowels are generally limited to closed syllables (read our blog posts here and here for more on the syllable as a useful notion for language learners). So for example the vowel of fem./plur. die is always long. Among these article forms, if the coda (final consonant) is voiceless, the vowel is short: this applies to das and des. The forms with voiced codas (i.e. all the rest) have long vowels.

1. Homophones and other patterns

If you struggle recalling the order of cases, note that they are in reverse alphabetical order: N-G-D-A (mnemonic: “Never Gonna Doubt Again!”). Note further that, everywhere except in the masculine, the outer two forms are the same. In the feminine, the inner two also match. In the masculine, pay attention to the sequence of those final letters: <r> precedes <s>, then <m> precedes <n> – just as it happens in the alphabet.

1. Basic vs replacement

One useful way to think about these 4 sets of words is that the masculine paradigm is basic, and all the others are derived from it by replacements. This lets us set up a rule that whenever we’ve replaced masc. nom. der, the accusative form matches it. Deriving the neuter from the masculine, we have just the one new form: nom./acc. das. Deriving the feminine and the plural sets, everything from the masc. paradigm is replaced, either by die or by der (note that fem. der is not nominative, unlike masc. der). For the plural paradigm, just remember RN, ‘registered nurse,’ for the endings of the inner 2 cases. Or, alternatively, the word “GaRDeN,” for the alignment Genitive–[r], Dative–[n].

Once you’ve successfully internalized these inflections, you are equipped to inflect any of the other determiner words! That is, with just one twist: weak final syllables of determiners contain no vowels other than /e/, so instead of die we have <-e> in dies-e, welch-e, solch-e; and instead of das we have <-es> in dies-es, welch-es, solch-es, etc. (and not *dies-ie or *dies-as).

2 Verb classes – Ablaut and Umlaut

First, let’s consider the basics of the principal ways English constructs the Preterite (past-tense) form of verbs. Three rhyming base forms nicely illustrate these diverse patterns: “sing,” “bring,” and “ping.” 

In the first of these, we see that the vowel of the base form is replaced: sing – sang. This curious strategy, it turns out, is an extremely archaic feature of our verb system. This vowel-replacement alternation is known by the German name “Ablaut” (or “apophony”). Note that, unlike the other two types below, no suffix is added to the base: it is the vowel change alone[2] which expresses the grammatical category Preterite. (In traditional Germanic grammar, such verbs are called “strong verbs,” as they don’t require the ‘help’ of a suffix.)

Next, we have the example bring – brought. Here we do see the addition of a suffix, /t/ in this case (sometimes /d/), as well as mutation of the stem. Because a suffix is added, this class and the next are traditionally called “weak verbs.” Often we find a shortening[3] of the long vowel of the base form (keep – kept, lose – lost), or reduction of a diphthong (hear – heard). 

Finally, we have the class commonly called “regular,” in which /d/ is suffixed (spelled <-ed> or <-d>), and the stem itself is not altered in any way: ping – pinged. Virtually all new verbs that enter the language are assigned to this class: google – googled, retweet – retweeted.

German, being a close sister to English, shows patterns very similar to those we see in our language. German for ‘sing’ is singen – sang (participle gesungen), and ‘bring’ is bringen – brachte (participle gebracht). As with the English verb, singen displays Ablaut – replacement of the root vowel, making this a “strong” verb. The Preterite brach-te, on the other hand, has an overt suffix – it is therefore a weak verb.

Now, both English and German exhibit another type of vowel alternation called Umlaut, or vowel mutation. This is distinct from Ablaut in being more tightly constrained: e.g. /a/ mutates only to /ɛ/, as in Eng. “man – men” or Ger. Mann – Männer. In German, Umlaut may refer either to the alternation pattern itself, or to the diacritic mark shared by the symbols <ä ö ü>, which represent the outputs of the Umlaut process.

Most of the time, when a German verb has a cognate with English, both will be strong, or both will be weak. (Exceptions exist: helfen – half ‘help’ and schlafen – schlief ‘sleep’ are strong verbs, while their English cognates are weak.) But there’s a fairly reliable rule that I hope will help you in the task of mastering German Preterite formation: If the base form of a verb has an Umlauted vowel, that verb does not undergo Ablaut. So, for a verb like spüren ‘sense, perceive,’ we find the Preterite spür-te with the suffix <-te>, and gähnen ‘yawn’ has gähn-te.

Like any worthwhile rule in grammar, this one has a few exceptions! These include:

• betrügen ‘betray’ – Pret. betrog

• lügen ‘lie, utter a falsehood’ – log

• hängen ‘hang’ (intransitive) – hing

• An important set of exceptions involve auxiliary verbs:

• dürfen ‘may’ – durfte

• können ‘be able’ – konnte

So keep your eyes peeled. Happy hunting! Below are some of the most frequent strong verbs in German, sorted first by vowel alternation, then alphabetically.

The mark ` indicates the onset of the root syllable, which bears stress. Note that the vowels of prefixes are excluded from variation. 

* These particular verbs are weak – they add a dental suffix – but I include them here because the root vowel varies.

† Like English “hold–held,” halten–hielt has a unique vowel alternation pattern due to its origin in reduplication: the Preterite reflects prehistoric *hé-hald- or *hí-hald-, one of the few remaining traces of a previously widespread process. The length of the vowel in hielt (before a consonant cluster!) is a further oddity resulting from this.

3 Constraints on word order

A feature of German that learners often struggle to fully master is its syntax. This term refers to the principles regarding the order of phrases (noun phrases, inflected verbs, etc.) making up a sentence. In some respects German shows greater diversity in its word order than English does – if the object precedes the verb in English, the result tends to sound like Yoda, and is not a way we should encourage English learners to talk. In German, however, putting the object first is just fine, provided it’s in a main clause (rather than a dependent one).

I have good news for you, dear learner: If you start with the right basic word order, you can derive all the permissible alternate orders with the application of just two simple rules!

Allow me to demonstrate. Here’s how German would express the sentiment ‘Today I will master German word order’ in a dependent clause:

Ich verkünde, dass | ich heute die deutsche Wortstellung meistern werde

The first string of words simply means ‘I announce that …’ – this type of framing sets up the word order we must regard as basic in the application of the two rules. Now, let’s set that aside and group the core proposition into phrases. I will mark Noun phrases with square brackets, and the Verb phrase with angled ones as in Word String (1).  

1 … [ich] <heute [die deutsche Wortstellung] meistern werde>

The subject, here ich ‘I’ (nominative), is a Noun phrase that is not part of a Verb Phrase. After a complementizer[4] like dass (or wenn, falls, weil, etc.), the subject must precede all other elements. The Verb phrase here contains one inflected verb, werde (which agrees with the subject, and must be final!), one verb in the infinitive, meistern; and aside from these it contains an Object Noun phrase [die deutsche Wortstellung] (accusative), and an adverb, heute. Note that the object tends to be closer than adverbs to the ‘end-zone,’ where all the verb words congregate.

Now let’s meet Rule A: “Front the inflected verb.” There can only be one such verb word in any clause; “front” simply means ‘move to the front.’ Everything else will remain in the same order. Applying Rule A to word order 1 yields word order 2, which is the one we find in Yes/No questions (not the only function, but a major one). The pipe | marks the boundary between the fronted word and the remainder in Word String (2).

2 Werde | [ich] heute [die deutsche Wortstellung] meistern?

Once this rule has applied, the Verb phrase is no longer contiguous as it had been previously – it has become disjointed by the operation of Rule A.

It’s time to learn Rule B: “Front any other phrase.” This rule applies to String 2 (not to String 1!) and is less specific than Rule A, and as a result produces multiple outputs (if the original clause has more than one candidate phrase). I will therefore label these “3a,” “3b” and “3c.” These three variants all express the same core proposition, but with differing points of emphasis. Remember that only one phrase (or constituent) is fronted, and all the rest remains as it was in String 2.

3a [Ich] | <werde heute [die deutsche Wortstellung] meistern>.

3b Heute | werde [ich] <[die deutsche Wortstellung] meistern>.

3c [Die deutsche Wortstellung] | werde [ich] heute meistern.

That’s it – these are the only orders of these words that sound good to German speakers. String 3a is obtained by fronting the Subject. This is the default candidate to be fronted. When anything else is selected, that constituent is being highlighted or foregrounded. In String 3b it was the adverb ‘today’ that was selected, and in 3c it was the Object.

Very important: the inflected verb (here werde) must stand in Position 2 in a declarative main clause! (That is not the same as saying it is the second word of the sentence – see 3c.)  

All other possible orders of the seven words in our target sentence are ungrammatical, and will sound either awkward, or downright awful! I recommend applying this type of derivation (mechanical application of rules to properly formed inputs) to various sentences you encounter. Start with simple ones, then work your way up to more complex clauses. With a little practice, this will start to seem very natural, and become automatic. If you apply the rules correctly, you will soon produce only correct word orders, and neatly sidestep all the ungrammatical ones.

CONCLUSION

So that about wraps it up for this post. We learned about two types of inflection: nominal inflection (or “declension,” Deklination) of articles and other determiners, and about verb inflection (or “conjugation,” Konjugation), specifically the formation of the Preterite, and the two broad types, called “strong” and “weak,” of Germanic verbs. Finally, we turned our attention to the level of syntax, and made use of rules in deriving multiple grammatically acceptable strings from a basic word order.

Let’s take another look at the definite article forms. In this table we’ve added color to draw attention to homophonous forms in neighboring cells. Being aware of patterns like this will help you double-check, when you’re practicing the paradigms, that you recalled everything correctly.

Last of all, I’d like to share some thoughts on the “why” of the exclusion of umlauted verbs from the Ablaut pattern of vowel replacement, mentioned in Section 2. 

Let’s suppose that the umlauted vowels /ä ö ü/ of German are in every instance derived, and that the rest of them /a e i o u/ are basic, or “primary.” Can we identify a consistent feature among the triggers, those elements which induce vowel mutation? Consider the shift from /a/ to /ɛ/ in the following pairs of related words in Table 4.

The type of mutation called Umlaut involves shifting from a vowel at point A to a point B that is one step closer to the vowel /i/. So this is a natural type of anticipatory change when a suffix with /i/, such as <-lich> or <-isch>, is added. This accounts for the mutation in täglich and isländisch, and thousands of other such derived words. 

But this change occurs not only with suffixes containing /i/! We find that the modern diminutive endings <-chen> and <-lein> also reliably trigger Umlaut. If we look at older forms of the language, or at even older reconstructed forms, we find that modern <-chen> descends from a suffix of the shape *-xin(n), which did contain /i/! And further, modern <-lein> derives from *līn, which had a long high front vowel. Umlaut in the Present 2s and 3s of verbs, finally, is well accounted for if modern <-st> and <-t> derive from older endings of the shapes *-ist, *-it/-id, but whose vowel /i/ was lost along the way.

Let’s be bold and state that all modern Umlaut vowels are traceable to pre-forms that once had /i/ in the following syllable[5]. In some suffixes this /i/ persists unchanged, as we see with <-lich> and <-isch>. In others, this /i/, short or long, has shifted to other vowels, and in yet another sit this /i/ was lost. The effects of this old vowel, however, remain.

What does that mean for the verb system? If we accept that the vowels /ü ä/ of spüren and gähnen mentioned above, for example, are derived, we would want to propose pre-forms such as *spuri-an and *gani-an, respectively – verbs that used to have a 2-syllable base. The presence of /i/ at that stage triggered Umlaut, but the trigger vowel itself was later lost. 

Now we can see a reason that such verbs are (mostly) excluded from the Ablaut pattern of vowel replacement: this replacement rule only worked well if the stem had a single vowel – otherwise confusion may arise regarding which vowel to replace! Verb stems with more than one vowel would therefore require some other means to form their Preterites, and the “weak” suffixing process does the job instead.

Once again, we hope this blog post is of interest and of use to you, no matter where you are in your learning journey. Happy studying – viel Erfolg! 

FOOTNOTES

[1] On cases: Here are the primary functions of the 4 cases of Modern German.

• NOMINATIVE is the case used for the subject of a sentence. English and German nominative pronoun forms include:

• GENITIVE marks a noun as possessor of another noun phrase. E.g. in Haus des Vaters ‘the father’s house,’ the possessor noun phrase des Vaters is in the genitive.

• DATIVE is the case of the person (or rarely thing) to whom something is given, conveyed, or related; dative also marks the experiencer in sentences like Mir ist kalt ‘I am/feel cold,’ lit. “it is cold to me.” 

• ACCUSATIVE is the case of the direct object, i.e. the entity most affected by a verbal action. English and German accusative pronouns include:

There is of course more to the story! More on cases in an upcoming blog post.

[2] “beat – beat” appears to be an exception; in Old English these forms had distinct vowels, which have since converged.

[3] In verbs of this class whose base form has a short vowel, the Preterite is often identical to the base: hit – hit, set – set, put – put. The (presumed) dental suffix (/d/ or /t/) here is fused to the dental consonant of the stem (/hit + t/ → /hit/).

[4] A complementizer is a functional category of words that can turn a clause into the subject or object of a sentence, e.g., words like “that” or “who” in English. 

[5] Confirmation is seen in old Germanic loans into older Finnish, such as modern kaunis ‘beautiful’ which is traced to *skauniz, also the source of German schön.