• Dutch
  • Frisian
  • Afrikaans
Show full table of contents
Historical and dialectgeographical information about the two IPI-constructions

The two types of Imperativus-pro-Infinitivo (IPI) differ with respect to their history and their dialect-geographical spread. The adjunct IPI is an old construction, going back to late Old Frisian, whereas the argument IPI is a recent development. In agreement with the preceding, the adjunct IPI is found in a much larger area than the argument IPI. The adjunct IPI is found in Dutch dialects and Low German, but it always has regular infinitival morphology, not the zero-morphology that is homophonous with the imperative, as in Frisian. Zero-morphology, or 'imperative' morphology, developed rather late in Frisian, in the 18-19th centuries, which is during the same period that the argument IPI began to be used.


The adjunct IPI originated in late Old Frisian, in the 14-15th centuries. This was the time when the subjunctive became morphologically indistinguishable from the infinitive, and subsequently got lost. The IPI originally had regular infinitival morphology. In some instance, the adjunct IPI could also be construed as an argument IPI, and, in many instances, the infinitival form of the verb was homophonous to the imperative form of the verb. This homophony was strongly reinforced by the declension of -je-verbs. These verbs systematically feature homophony of the imperative and the ordinary infinitive throughout the history of Frisian from at least the 16th century onwards.

How the process took place exactly is not known, but the eighteenth and nineteenth centuries saw the genesis of two simultaneous developments:

  1. The form of the verb that was homophonous to the imperative came to be used in the IPI instead of the infinitival form (normally ending in schwa).
  2. The argument IPI came to be used, whereas the adjunct IPI was being used less.

There may or may not be a connection between these two developments. In the 20th century, there were speakers who used the argument IPI but not the adjunct IPI. If these speakers were mother tongue speakers of Frisian, then the IPI was also transferred into their Dutch.

The adjunct IPI-construction is by no means restricted to Frisian. It is also found in Dutch coastal dialects and Low German dialects and in seventeenth-century Dutch. There is, however, a difference with present-day Frisian. In Dutch and Low German dialects, the adjunct IPI always, to the best of our knowledge, features a verb which bears overt infinitival morphology. Recall that the Modern Frisian infinitive in this very construction featured a form of (zero) infinitival morphology that is homophonous to imperatives.

The argument IPI is restricted to Frisian. It is practically certain that it does not occur in dialects of Dutch, not even in those featuring the adjunct IPI. It is probably the case that the argument IPI does not occur in Low German either, but this language variety has not been well researched syntactically.

The table below summarises in which West Germanic varieties which type of Verb-Initial infinitival may be found:

Table 1
Adjunct IPI Argument IPI Inflection less infinitive
Modern Frisian Yes Yes Yes
Middle Frisian Yes Around 1800 Around 1800
Dutch dialects + Low German Yes No No

It can be understood why in Dutch dialects the infinitive (inside the IPI) did not develop into an imperative. Middle Frisian had this development due to its regular homophony of imperative and infinitive, for which the JE-class of verbs was responsible. The dialects of Dutch, however, did not feature any declension of JE-verbs, and hence there was no basis in Dutch for the subsequent development of the IPI which we know from late Middle Frisian. Hence it comes as no surprise that Dutch dialects featuring the IPI only only display the infinitive inside it, not the form which is homophonous to the imperative.