German verb conjugation described precisely with a mechanical tool
by Stephen Hewitt | Published September 2006
- The essentials of German verb conjugation
- The endings of the verbs and their use
- Formation of the appropriate stem
- Adding the endings to the stem
- Irregularities and the information necessary to conjugate a verb
- The use of gremple
- Further work
Gremple is a German verb conjugation tool. Its purpose to help non-native speakers to learn to conjugate German verbs. The idea underlying Gremple is to develop a precise description of German verb conjugation, with a minimum of data stored for each verb, and as much of the work of conjugation as reasonably possible transferred into a system of rules. This aids learning because it is easier to understand a system of rules, particularly where their underlying motivation is reasonably apparent, than to memorise a large quantity of apparently disordered (and therefore meaningless) data.
As a learning aid, Gremple provides:
- a reliable reference to the conjugation of many German verbs
- a precise explanation of how a particular verb form is derived, through the rigid application of the set of rules described here.
- a concise summary of the irregularities of each verb, in other words the information necessary to conjugate the verb.
In addition to introducing Gremple, this document provides a precise description of German verb conjugation. This description can be used completely independently of Gremple as any other description from a text book can be. However, a benefit of Gremple is that the rules of conjugation described here have been tested. These are the rules that are embedded in the code of gremple. This does not exclude altogether the possibility of error; it is possible that the rules as formulated here in natural language do not match the rules encoded in the programme, for example, but it does introduce some rigour. The existence of a mechanical tool enables generalisations about all verbs to be formulated and immediately tested, and the regularity of verbs to be demonstrated.
German verbs are in general regular, more regular than the popular textbooks allow. For example Hammer on page 238 mentions "the wholly irregular verbs, i.e. the modal auxiliaries and wissen, and the verbs sein, haben and werden". In fact, apart from sein, these verbs are far from being "wholly" irregular. Some of them are almost completely regular, as we will see. Gremple's concise summary of the information needed to conjugate each verb makes this clear.
German verbs are classified here as either strong or weak. By definition a strong verb is one in which the the stem of the past tense differs from the stem of the infinitive. For example treten has the past tense ich trat. The e of the infinitive has become an "a". Every verb that is not strong is weak. Note that the definition here is slightly different to the traditional one, which holds that the vowel of the past stem must differ in order for the verb to qualify as strong. The only verb whose classification is affected by this difference in definition is haben. This allows us to catch haben as a more or less regular strong verb with weak endings.
Unfortunately the study of the conjugation of German verbs has been made more confusing than it need be by the misuse of the word "regular". It is traditional and widespread to use "regular" to mean weak, and "irregular" to mean strong. This might not matter - we could after all call them oranges and lemons if everyone agreed on that - were it not for the fact that "regular" and "irregular" have normal, non-specialist meanings in English which are more or less indispensable in the discussion of German verb conjugation. These words will be used in this text in their standard English meaning. Hence "regular" will imply a verb which conjugates in an ordered way, according to a set of well-defined rules. However, it is worth noting that "regular" and "irregular", rather than representing a binary choice between black and white, are used here like points on a scale of greyness.
Only the four simple tenses of the verb and the so-called "past participle" Partizip II are discussed here. The compound tenses are excluded since they are always regular and the same for all verbs. The simple Konjunktiv I tense is abbreviated "K1" and the simple Konjunktiv II as "K2". The terms "personal irregularity" and "stem irregularity" are defined in the section Irregularities and the information necessary to conjugate a verb below. Kranton stems are defined in the section Adding the endings to the stem
Rather than writing things like "third person plural", each different form of a tense will be identified by its pronoun. The pronouns are shown in the following table.
|ich||1st person singular|
|du||2nd person singular|
|es||3rd person singular|
|wir||1st person plural|
|ihr||2nd person plural|
|sie||3rd person plural|
The essentials of German verb conjugation
1. In all tenses for all verbs without exception the sie form (3rd person plural) is identical to the wir form (1st person plural). For this reason, no more information will be presented here on the sie form. Note that the so-called "polite" Sie form is identical to the 3rd person plural sie form.
2. The present tense wir form is identical to the infinitive of the verb, with the sole exception of sein, where it is "sind".
With the exception of the present tense wir form (and consequently, by rule 1, the sie form too) conjugation of each tense consists essentially in adding the endings tabulated below onto the appropriate stem of the verb. To be precise, regular conjugation consists of adding the endings while applying the phonetic rules given below. The definition of the appropriate stem is also in a section below.
Formation of a regular Partizip II consists of adding the ending to the stem and applying some rules to determine whether it has a prefix ge-.
The endings of the verbs and their use
|present||weak past||strong past||Konjunktiv|
|Partizip II||-t||infinitive ending|
The present endings in the above table are used in all verbs. The "strong past" endings are used in regular strong verbs. The "weak past" endings are used in all weak verbs. The Konjunktiv endings are used for the K1 of all verbs. They are also used for the K2 of regular strong verbs. Weak verbs use the "weak past" endings to form K2. (This means that for every weak verb K2 and past are identical)
There exists a sub-class of strong verbs which exhibit a particular irregularity. By definition a verb in this sub-class exhibits the particular irregularity that it takes the same endings (in every tense and in Partizip II) as would a weak verb. By tradition there are only nine such "mixed" verbs, namely brennen, bringen, denken, kennen, nennen, rennen, senden, wenden and wissen. That tradition is misleading because in the same sub-class are in fact also dürfen, haben, können, mögen and müssen.
The definition of the "infinitive ending" in the table is that part of the infinitive which is left after the infinitive stem, defined below, is taken away. The infinitive ending will always be either -n or -en.
Formation of the appropriate stem
With the sole exception of knien (where it is knie), the stem of the verb is derived by removing an -en from the end of the verb if possible, and otherwise by removing the -n (RULE1). Every verb ends in -n. For example, the stem of tun is tu, the stem of sein is sei, the stem of handeln is handel and the stem of machen is mach. Let us call this stem the infinitive stem.
What is left of the infinitive after this stem has been removed is the infinitive ending, which is used as the ending of Partizip II of a strong verb. For example the infinitive ending of tun is -n and the infinitive ending of lassen is -en.
For weak verbs, the infinitive stem is used to form all the tenses, except that with -ern and -eln verbs, there is a possibility of dropping the last "e" in the stem from the ich present and the ich and es of K1.
In general there are no rules which predict from its morphology whether a verb is weak. However, all verbs ending in -eln, -ern, and -lichen are weak. All ieren verbs are weak, if we define an ieren verb to be one that not only ends in -ieren, but also has other syllables to which the -ieren is a sort of suffix. For example frieren is not an ieren verb under our definition because if we take away the -ieren, there are no syllables left. Incidentally ieren verbs under this definition form Partizip II without a "ge-" prefix. Similarly verbs ending in -igen not immediately preceded by a vowel are all weak. For example einigen is weak but schweigen ends in -igen immediately preceeded by another vowel and is strong.
For regular strong verbs, the infinitive stem is always used for K1 and the ich, wir and ihr forms of the present tense. However, there are four other specific stems.
- One stem is used for the past tense.
- One stem is used for Partizip II.
- One stem is used for K2.
- One stem is used for the du and es forms of the present tense.
Although for any particular verb some of these stems will actually be identical to each other, and some will be the same as the infinitive stem, they have the potential in general to be different to each other. In other words, if we took any pair of the stem types from this list, there would exist a least one verb for which those two stems would be different to each other.
In the more regular strong verbs, and in the majority of strong verbs, only the vowels differ between the different stems. In slightly less regular verbs, the consonants on the end of the stem may also differ between the different stems, but the consonant changes are much more constrained than the vowel changes. Without exception, for a given verb, only one alternative set of consonants is allowed between all the stems for all the tenses. In other words, if one of the stems has alternative consonants to those in the infinitive, then the consonants of every other stem must either have those consonants, or the same consonants as the infinitive. For example, nehmen has a stem with a double m in the du and es forms of the present tense (du nimmst), and the same alternative consonants are also used in the Partizip II (genommen), while all its other stems retain the same consonants ("hm") as the infinitive. By contrast the vowel is different in every stem, changing from "e" in the infinitive to "i" in the du and es of the present tense, to "a" in the past, to "o" in Partizip II and to "ä" in K2.
In the highly irregular sein, the whole stem differs - the infinitive stem is "sei", but the past stem is "war".
A completely regular strong verb will form its K2 stem by applying an umlaut to the vowels in past stem if the vowels are capable of taking an umlaut, or using the past stem unchanged otherwise (RULE2). Even for verbs with an irregular K2 stem, the consonants in the K2 stem are always the same as the consonants in the past stem. Without exception, the K2 stem differs (if at all) from the past stem only in the vowels.
Apart from the K2 stem, the question of whether it is possible to predict some or all of the stems from the morphology of the verb is beyond the scope of this current article and has not yet been tested with gremple. This is an area for possible further work. As far as gremple is concerned, the information necessary to form these stems is taken as given, ie data that must be stored for each verb and this corresponds to what a learner needs to memorise.
Finally for strong verbs, we also note that it is relatively unusual for the consonants in the stem of the past tense and in Partizip II to differ. Apart from the highly irregular sein, the only examples are nehmen, where the Partizip (genommen) has alternative consonants and the past stem (nahm) does not, and backen, where the past stem has alternative consonants (buk) and the Partizip (gebacken) does not.
For clarity, it is worth affirming explicitly what the earlier definition of a strong verb and our assertions about weak verbs together necessarily imply. A strong verb is defined as one which has the past tense stem different to the infinitive stem. Yet every weak verb uses its infinitive stem in every tense. This means that there are no German verbs that have the past stem and the infinitive stem the same and yet (for example) the K2 and Partizip II stems different. If the past stem is the same as the infinitive stem, then all the stems must be the same, and the verb is called weak.
Adding the endings to the stem
The definition of a kranton stem is this. All stems ending in -d and -t are kranton stems. All stems ending in -n and -m are kranton stems, unless the last three letters in the stem match a pattern in one of the rows of the following table. No other stems are kranton stems.
|last but 2||last but one||last letter|
|(doesn't matter)||vowel||"m" or "n"|
|vowel||"l", "r" or "h"||"m" or "n"|
Kranton stems require special treatment with any ending that starts with a consonant. Note that such endings are only the -st and -t of the present tense and strong past and all weak past endings.
Except for the four cases listed below, a padding "e" is always inserted between a kranton stem and any ending that starts with a consonant. The exceptions are:
- The forms of a verb where a present tense vowel shift is operative. This means the present du and es forms of a strong verb which has a present tense vowel shift. In this case, no padding is used, but the following rule is applied instead. If the stem already ends in the ending in question, then the ending is simply left off, or, in other words, becomes zero. Otherwise the ending is appended as normal.
- In Partizip II of strong verbs with weak endings or the past tense of strong verbs with weak endings, no padding is ever inserted. (Examples are haben, senden, wenden)
- In the k2 of strong verbs with weak endings padding is inserted if the k2 stem is the same as the infinitive stem (for example with wenden, senden it is inserted, but with haben it is not).
- In the du form of the past of a strong verb, the padding is optional.
Examples of conjugations with kranton stems
The table demonstrates the difference in padding between verbs with present tense vowel change and those without. Considering first the du form, bersten has a present tense vowel shift, so that its stem for the es and du forms is birst. Now the ending for the du form is -st, which is already on the end of the stem, so the ending is simply omitted under rule 1 above. By contrast, treten, although it has a present tense vowel shift, does not have a stem ending in -st, so the -st present du ending is appended under rule 1 with no padding. In the es form, the both testen and bersten already have the ending (-t) on the end of the stem, so it is omitted in both cases.
Finally two other verbs are shown, the strong bitten, which has no present tense vowel shift, and the weak verb testen. For both bitten and testen, because no present tense vowel shift is operative, a padding "e" is inserted between the kranton stem and the ending, in both the du and es forms.
In the present tense, sibilant stem endings that cause a leading s from the ending to be dropped are -x -s -ß and -z.
In the past tense, the same sibilant endings and also -sch cause a padding "e" to inserted between the stem and the ending that has a leading s. (Note that the only past ending that has a leading s is the du form of a strong verb.)
Irregularities and the information necessary to conjugate a verb
The more irregular a verb, the more information is needed to conjugate it. For a regular weak verb, we would need nothing more than the infinitive and the knowledge that it was regular and weak. For a regular strong verb, we would need the infinitive, the knowledge that it was regular and strong, and the vowels needed to form the past stem, the Partizip II stem, the knowledge of whether there is a present tense vowel shift in the du and es forms, and if there is, what the vowel is.
With the exception of knien and the modal verbs sollen and wollen, all weak verbs are entirely regular. Strong verbs have varying degrees of irregularity.
Now, with the exception of the present tense of sein, irregular verbs do not exhibit unconstrained, arbitrary irregularities. There are only certain types of irregularities that a verb may have.
One of these has already been mentioned: a strong verb may take weak endings. Note that these are not mixed arbitrarily with strong endings; for any particular strong verb, either all the endings in every tense and Partizip II are weak, or none of them are.
Note that there are some verbs, such as fragen, that have more than one way of being conjugated, in some or all of the tenses. In the case of fragen for example, usually it is a regular weak verb, but it can also be conjugated in the past as ich frug etc. With many of these verbs both ways of conjugating the verb in question are regular. For example for fragen, the conjugation frug etc is a regular past strong conjugation and the more usual weak conjugation is also completely regular. However the fact of having both possibilities must be considered itself an irregularity; at any rate it is something that must be learned for each verb. The constraints formulated here are not intended to apply to all conjugations of such verbs in their entirety but only to the one way of conjugating them considered at a time. In other words, we can apply these rules to fragen as a weak verb, forgetting for a moment that it can be conjugated as a strong verb in the past tense. Then we can apply the rules to conjugate the past tense of fragen, forgetting that it can be conjugated as a weak verb. But for example, it would be wrong to apply the rule that says either all the endings of a verb are weak, or they are all strong, to both the weak and strong version of fragen because clearly if this rule were applied in this manner, it would be shown to be incorrect.
For the remaining kinds of irregularities, it is necessary to make an important distinction. This distinction is between irregularities in the formation of the stem for a particular tense and irregularities in individual finite forms within a given tense. For example, sein has a highly irregular past tense stem "war" but, given the stem, the past conjugation is completely regular. We will call this kind of irregularity a stem irregularity. By contrast, the K1 of sein has a regular stem "sei" but some of the finite individual K1 forms (ich, du and es to be precise) are irregular. We will call this kind of irregularity a personal irregularity.
Now, most irregularities occur in the formation of the stems. Given the stem for a particular tense and the fact of whether the verb takes weak endings, with very few exceptions, the conjugation of the tense is entirely predictable. Such personal irregularities as do occur are mainly in the present tense of less than a dozen verbs. Exact details for each tense are presented under the headings below. Here we note some universal constraints on personal irregularities.
1. With the sole exception of the present tense of sein, personal irregularities occur only in the singular.
2. Also with the sole exception of the present tense of sein, if there is a personal irregularity in a tense, then the ich form always becomes identical to the es form.
Present tense irregularities
The present tense has more irregularities than the other tenses. However, even in the present tense, the irregularities are quite highly constrained. Firstly we note that, with the exception of sein, all irregularities are confined to the singular. Remember that our definition of a regular strong verb allows it to have a different stem in the du and es forms of the present tense, so we do not consider verbs such as treten (with ich trete du trittst er tritt) to be irregular.
Given this definition, only a handful of verbs have irregularities in the present tense. These are sein, werden, wissen and the modal verbs dürfen, können, mögen, müssen, sollen and wollen
Firstly, the present tense of sein does show unconstrained irregularity, and it will appear as an exception to all the constraints formulated here for the present tense. However note that even here, the universal rule (2) that the sie form is identical to the wir form, still applies.
Given that it is a strong verb with weak endings and past stem "hat", haben is conjugated in a regular way except for the present tense "du hast" and "es hat". However, if we allow it to have a different present stem "ha" for du and es forms (which regular strong verbs are allowed) we can consider it to be entirely regular (given the initial assumptions above).
In other words, we are saying that haben is an entirely regular conjugation of a strong verb with weak endings, past stem "hat" and present du and es stem "ha". Those three pieces of information - the presence of weak endings and the actual two stems - are all that is necessary to completely conjugate haben in all the tenses.
With the exception of sein and werden, there are only two possible personal irregularities which can occur in any verb in the present tense.
1.The dropping of the endings from both the ich and es forms. Note that the irregularity consists in dropping both of them - either they both must be dropped, or neither of them. Examples of this are all the modal verbs and wissen. Furthermore, if the endings are dropped, then the stem in the singular present tense must be the same for all persons. A strong verb which drops these endings cannot also avail itself of the right, which regular strong verbs have, of having a different stem for the du and es forms but not the ich. We note as a consequence of these observations that in all cases of this irregularity the ich and es forms will be identical (consisting solely of the stem). (Most verbs that exhibit this irregularity also exhibit the next one; sollen is the sole exception that prevents this being formulated as a general rule.)
2. The other possible personal irregularity is a stem which is different to the infinitive stem throughout the singular of the present tense (rather than in just the du and es forms as is possible for a regular strong verb). This stem is constrained to differ only in its vowels from the infinitive stem. In addition, any verb which exhibits this irregularity must also exhibit the irregularity 1. Examples of this are wissen and all the modal verbs. Note that wollen has this irregularity even though it is weak.
Past tense irregularities
Only werden displays personal irregularities. As for stem irregularities, many strong verbs change the trailing consonants but only sein changes the leading ones too.
The K1 stem is always regular (it is the stem from the infinitive). The only verb with personal irregularities is sein.
There are no personal irregularities with any verb. All weak verbs have regular stems too, and consequently are completely regular. There are many strong verbs have an irregular stem. This means a stem where the vowel is not formed according to RULE2 above. But, to re-iterate, the irregularity occurs only in the vowels. Without exception the K2 stem differs (if at all) from the past stem only in the vowels.
Partizip II irregularities
We have already stated that strong verbs may have the irregularity of weak endings. Of course this will result in a weak ending on the Partizip. This is not considered here, since it is not classed as a Partizip II irregularity, but a more general one. To re-iterate, it is not possible to have a Partizip II with a weak ending unless the verb also has weak endings in every tense.
The reverse is not true, however. Certain irregular weak verbs have Partizip II with strong endings, for example mahlen, erschallen and the weak version of backen.
There are a certain very few other verbs with minor irregularities in Partizip II. For example essen has gegessen where the regular form would be *geessen and schreien can have either the regular geschrieen or geschrien, so similarly can speien.
Partizip II of sein demonstrates arbitrary irregularity. werden loses the ge- when used as an auxiliary.
The modal verbs present something of a philosophical dilemma. When used with a bare infinitive, the normal Partizip II of a modal verb does not occur, but instead something that looks identical to the infinitive of the modal is used. For example, "ich habe es nicht machen können". Now the dilemma is, is the können" in this example - which syntactically is playing the role of a participle - is it really a different form of the participle which just happens to look and sound identical to the infinitive, or is it really the infinitive itself? Both Hammer (page 261) and Helbig and Buscha (page 109) seem clear that it is the infinitive itself. If so, we do not have to consider it as an irregularity of Partizip II (because it isn't Partizip II). That is the approach adopted here. Further support for this idea is that the same can occur with lassen, sehen and hören We should note, however, that verb conjugation books (eg Strutz) do list it as an irregularity of the Partizip II for the modal verbs (but, perhaps inconsistently, not for lassen, hören and sehen).
The use of gremple
As noted in the introduction, gremple stores the minimum amount of information for each verb. When a user wants to see the derivation of a verb conjugation, gremple presents this information and the steps by which the final conjugated forms are derived from it. As explained above, the information stored by Gremple corresponds to what a learner needs to know to conjugate the verb.
For strong verbs, information about the stems is stored for each verb, with the exception of a regular K2 stem, which can be deduced using the rules noted above. To conjugate a verb, gremple adds the endings described above under The endings of the verbs and their use using the system of rules described under Adding the endings to the stem. Irregular stems are simply specified in the information given for the stems. Weak endings on a strong verb are specified by a special flag ("-w").
For personal irregularities, gremple has a simple notation which allows individual finite forms of a verb to be specified, over-riding whatever form would have been produced by application of the regular rules. This means that currently gremple does not capture any of the constraints on personal irregularities described above. For example, we noted that with the sole exception of the present tense of sein, if there is a personal irregularity in a tense, then the ich form always becomes identical to the es form. Gremple is oblivious to this, and requires both the ich and es forms to be independently specified for a verb if they are irregular. The same system is used for Partizip II. The irregular participle must be specified in full, and there is no notation for example, to say something like "although this is a weak verb, the participle should have a strong ending". This might be regarded as a limitation of gremple, but this was chosen for gremple's notation because it is intuitive and easy to understand and because the number of cases where the missing rules apply is very small.
The gremple notation for specifying how a verb is conjugated
The following table lists every possible irregularity and its corresponding representation (or lack of representation) in the notation used for the data stored for each verb in Gremple.
|Irregularity||Its representation in Gremple|
|The stem used in the present tense, and in the case of a weak verb for all tenses, is not formed according to RULE1 above. The only verb with this irregularity is knien.||STEM keyword to specify what the stem used in all the tenses is.|
|A strong verb uses weak endings in all its tenses and in Partizip II||-w flag.|
|A strong verb forms the vowel in its K2 stem other than by the RULE2 above.||The K2 vowel is specified either as the fourth string of letters after the headword, or following the keyword k2|
|A strong verb has a different stem for the du and es forms of the present tense||The vowel (or vowel and following consonants) are specified in the third string of letters after the headword.|
|A verb has a stem in all singular forms of the present tense which is different to the infinitive stem.||This irregularity cannot be captured by gremple. Instead all three singular forms of the present tense must be individually specified as if they were arbitrary personal irregularities.|
|A strong verb has consonants on the end of one or more of the du-es present stem, past stem and Partizip II stem, which are different to the consonants on the infinitive stem.||For each of these stems the consonants can be specified by appending them to the vowels specified for that stem (see below).|
|A verb drops the endings on the ich and es forms of the present tense and has a different stem for all singular forms of the present tenses.||This irregularity cannot be captured by gremple. Instead all three singular forms of the present tense must be individually specified as if they were arbitrary personal irregularities.|
|A weak verb has Partizip II with a strong -en ending.||This irregularity cannot be captured by gremple. Instead the whole of Partizip II must be specified as if it had an arbitrary irregularity.|
The notation for the stored information consists firstly of the infinitive of the verb. If the verb is weak and regular, then no more information is necessary. For example, the entry for bilden is:
If the verb is a strong verb, then the infinitive is followed by a string of lower case letters specifying the stem for the past tense. For this string, there are three possibilities:
- This string can consist of vowels only, in which case it means that the past stem is the infinitive stem with just the vowels substituted.
- It can consist of vowels followed by consonants, in which case it means that the past stem consists of the infinitive stem with both the vowels and the trailing consonants replaced by those in the string.
- Finally, it can start with a consonant, which means that the string completely replaces the infinitive stem. This happens with sein, where the infinitive stem is replaced by "war" in the past.
The string specifying the stem for the past tense is followed by similar but optional strings specifying, in this order, the stem for Partizip II, the stem for the du and es forms of the present tense and finally the stem for K2. This intuitive notation is similar to that adopted by some verb table books (eg "German Verbs", William Rowlinson, Oxford University Press, 1994). If it is necessary to specify a string later in this list without specifying one of the earlier ones, then the earlier ones can be replaced by a "-" as necessary. For example:
sehen a - ie
This means that the stem for the past tense of sehen is sah, the stem for Partizip II is unchanged from the infinitive stem, and the stem for the du and es forms of the present tense is sieh. Finally because no K2 stem is specified, it means that the K2 stem is formed by the regular rule (see above). Applying an umlaut to the vowel in the past stem means that the K2 stem is "säh"
Alternatively, the K2 stem can be specified without necessarily having 4 strings by preceding its string with the keyword "k2". (It still has to be the last string though). For example:
kennen a a k2 e -w
This means that the stem for the past tense is kann, the stem for Partizip II is also kann, and the stem for K2 is kenn. Since no stem is specified for the du and es forms of the present tense, this indicates that it is the same as the rest of the present tense. Incidentally, this example also illustrates the use of the keyword "-w" which means that although the verb is strong, it takes weak endings.
An example that illustrates changed consonants in the stems is nehmen:
nehmen a omm imm
This means that the stem for the past tense is nahm (with the vowel in the infinitive changed to an a) and the stem for the participle is nomm (with both vowel and trailing consonants changed from the infinitive) and similarly the stem for the du and es forms of the present tense is nimm.
The keyword PTP introduces an irregular Partizip II, for example:
essen a e i PTP gegessen S auf
The regular participle produced by the rules of conjugation would be geessen and this entry causes it to be replaced by the correct gegessen. Following the PTP, a "/" introduces an alternative form for the participle. For example:
schreien ie ie PTP/geschrien
This means that in addition to the regular geschrieen, there is an alternative form "geschrien".
The keywords PRESENT, PAST, K1 are used to introduce personal irregularities in each of those tenses. Note that there is a keyword K2 but it cannot be used in this manner because as noted above, there are no personal irregularities in K2. For example, after "PRESENT" would follow a list of irregular personal forms. These start with the ich form and by default increment forward through the conjugation, except that the keywords ich du es wir and ihr may be used to jump to a specified form. For example:
sollen PRESENT soll es soll
This means that the ich form of the present tense of sollen is "soll" and the es form is "soll". Apart from this, sollen is a regular weak verb. This is one of the modal verbs that Hammer misleadingly describes as "wholly irregular".
Other keywords used in the notation are associated with specifying alternative forms for some verbs. The "|" symbol introduces an alternative way of conjugating the verb that is treated like a complete and additional separate entry. The "+" symbol introduces a new specification of an alternative way of conjugating certain specified tenses. The syntax for an entry following "+" is the same as for the main entry, except that the symbol "'" may immediately follow the "+". The meaning of the "'" is a sort of ditto, and it indicates that all the stem information (but not the personal irregularities) is duplicated from the preceding entry. After a "+" entry, the keywords PRESENT, PAST, PTP, K1, K2 may be used to indicate which of these tenses are being specified as alternatives. All of them, except K2, can simultaneously be used to specify irregular forms as in the main entry. For example
The first entry (before the "+" sign) means that stecken is a regular weak verb. The alternative forms entry, (after the "+" sign) means that stecken has alternative forms as a strong verb, with a stem in the past tense of stak. The presence of the K2 keyword means that the K2 of this version of stecken is also a valid alternative form. Comments in square brackets  can be used to distinguish different contexts in which alternative forms of the verb conjugation are appropriate. For example here, the strong forms can only be used if the verb is intransitive.
Gremple's explanation of the derivation of the conjugations
For each tense, gremple displays a table containing the usual six personal forms. Partizip II has a similar table but with only one line. In the tables, the column headings have the following meanings.
- "stem" refers to the stem used to form that tense. It will always be the same for all rows in the table, except in the present tense where the du and es forms may have a different stem, as explained above.
- "padding" refers to the "e" that is sometimes inserted under the rules explained in Adding the endings to the stem.
- "standard ending" refers to the regular endings appropriate to the tense, as defined in the tables above in The endings of the verbs and their use.
- "modified ending" refers to the possibility of removing the leading s from an ending or removing the ending altogether according to the rules defined in Adding the endings to the stem.
- "irregular form" refers to a personal irregularity as defined above. This is an irregularity in an individual personal form of the tense, which has been simply memorised by gremple.
The final column of the table represents the actual conjugated form of the verb. As noted above (see The essentials of German verb conjugation) the regular wir form of the present tense is the infinitive and without exception the sie form is identical to the wir form in all tenses. In all other cases, if there is no "irregular form" in that row of the table, then the final column will be the concatenation of the stem followed by any padding followed by the ending. If there is an "irregular form" then that irregular form is the conjugated form of the verb.
Mentioned above is the question of whether it is possible to predict some or all of the stem changes in a strong verb merely from knowing the infinitive. There is no doubt that there is some regularity within the pattern of stem changes. There are in fact recurrent patterns in these changes and the learner ought to be familiar with them. At present Gremple cannot directly help with this, because it makes no attempt to capture generalisations about likely vowel changes. Instead it requires the changes to be completely specified for each verb in isolation.
There exist schemes for classifying strong verbs into different groups according to the vowel in the infinitive stem and how it changes.
Strutz arranges the verbs into 6 groups according to the vowel in the infinitive, with two subgroups in each group - each subgroup having the same pattern of vowel changes. However he gives no way of predicting which subgroup a given verb should go into. Helbig and Buscha have 8 groups (pages 37 - 40) most with two subgroups, and the advantage that assignment to a group and subgroup depends on the vowel in the infinitive and the letters that follow it. They write "Auf diese Weise kann man aus dem Vokal des Präsens und dem folgenden Laut darauf schließen, welcher Vokal im Präteritum und im Partizip II stehen muß."
There appears to be scope therefore, to reduce the data that gremple stores for a strong verb. It might be, for example, that most strong verbs need nothing except the knowledge that they are strong and then the stems can be predicted from the rules in Helbig and Buscha.
Both Gremple and this description of conjugation currently make no attempt to cover the imperative. Gremple gives no indication of which of the two auxiliary verbs haben and sein is used to conjugate the compound tenses.
Hammer. "Hammer's German grammar and usage", 2nd edition, Martin Durrell, Edward Arnold, 1991
Helbig and Buscha. "Deutsche Grammatik", Gerhard Helbig & Joachim Buscha, Langenscheidt, 1991
Strutz. "501 German verbs fully conjugated in all the tenses", 2nd edition, Henry Strutz, Barrons, 1990
- Vigilia por Julian Assange en la embajada de Ecuador Domingo 23 de septiembre 2018, Londres, 3 Hans Crescent, SW1X 0LS.
- Vigil for Julian Assange at the Ecuadorian Embassy 23 September 2018, London, 3 Hans Crescent SW1X 0LS
- Rosi on filming real-life confessions of a drug hitman in Mexican police 21 May 2017, Cambridge Arts Picturehouse, Regent Street
- How to remember a provably strong password: a new way using ‘constrained choice’ July 2018, Stephen Hewitt. The 2nd Clarion data privacy article