A shared feature of the Germanic languages is the occurrence of complex verbs consisting of the verb itself and what I refer to as the adverbal unit (AU). I examine the nature of the units that can be inserted into such complex verbs and compare intercalation patterns in AU-Vs and V-AUs. AU-Vs are found to be much more resistant to intercalation than V-AUs. The former accommodate the past participle marker, the infinitival linker, and—less commonly—verbs, whereas the latter accommodate NPs, ADVPs, and—less commonly—both phrase types concurrently. Thus, V-AUs may be split by more syntactic as well as heavier material than AU-Vs. I argue that this difference in cohesiveness is due to varying degrees of coactivation of Vs and AUs. The constituents of AU-Vs show a higher degree of coactivation than those of V-AUs. Adverbal units depend for their activation on the prior activation of verbs more than verbs depend for their activation on the prior activation of adverbal units. These different activation patterns lead to different degrees of cohesiveness and hence to different intercalation possibilities in the two verb types. Although intercalation is compulsory in some contexts, it proves to be a dispreferred option.*

This article reviews some of the main theoretical claims made in Jonathan David Bobaljik’s 2012 book, which deals with root suppletion in adjectival degree expressions. My first goal is to make the reader familiar with a coherent fragment of the overall system and the data that motivate it. The second goal is to discuss one specific part of the account, namely that words, understood as complex heads, are special for suppletion in the sense that suppletion is impossible beyond this domain. I argue that it is possible to abandon this assumption with no loss of descriptive coverage, and argue that in doing so, we can formulate a unified theory which covers both suppletion and morpheme order.