jats:titleAbstract</jats:title>jats:pThis paper investigates the development of sentential negation in Middle High German using sermons from the Upper German dialect area. To this end, a heterogeneous yet fine-grained corpus of Alemannic and Bavarian sermons is analysed with respect to diachronic development, geographical distribution and language-internal factors. What becomes clear is that Jespersen’s Cycle, a cross-linguistic model of the development of negation that can be seen as part of the received history of German negation, fails to account for the mechanisms in the development of sentential negation in German. These mechanisms cannot be understood independently of the – in some respects parallel – development of n‑indefinites. It appears that the interplay of variation and the grammaticalisation of the n‑indefinite<jats:italic toggle="yes">nicht</jats:italic>, which co-occurred with<jats:italic toggle="yes">ne</jats:italic>but could also appear on its own, played a more important role in the emergence of the negation particle<jats:italic toggle="yes">nicht</jats:italic>than previously thought. It is argued that when<jats:italic toggle="yes">nicht</jats:italic>was grammaticalised it retained the variation of the n-indefinite<jats:italic toggle="yes">nicht</jats:italic>, and that the subsequent loss of<jats:italic toggle="yes">ne</jats:italic>was a parallel development in both usages of<jats:italic toggle="yes">nicht</jats:italic>.<jats:fn symbol="1" id="j_bgsl-2017-0001_fn_001_w2aab2b8b1b1b7b1aab1c14b1c15Ab1">jats:pFür wertvolle Hinweise zu einer früheren Version dieses Beitrags möchte ich mich bei Stephan Elspaß, Helmut Graser, Nils Langer, Sonja Müller, Konstantin Niehaus, Hannes Scheutz und zwei anonymen GutachterInnen herzlich bedanken.</jats:p></jats:fn></jats:p>