It may be worthwhile to examine some of the intermediate results of the generation process. To keep the protocol as readable as possible some of the details have been omitted. A version with less omissions which gives more details about the annotated syntax graph formed by Constituent is in the Appendix.

Step 0:
(DescribeS I-EinführenKameratrokar :as :NP-elliptic)

Step 1:
(Sentences (DescribeS #
:BY t-consecutiveProcesses
:AS :NP-elliptic))

The original ToDescribe form is replaced by the one defined by the semantic function t-consecutiveProcesses.

Step 2:
(GOV-CONSTITUENT
   (AGGREGATION COREDAGGOBJ
      (DESCRIBES #
         :BY T-HANDLING :AS :NP)))
      (DESCRIBES #
         :BY T-HANDLING :AS :NP)))
      (DESCRIBES #
         :BY T-HANDLING :AS :NP))))
      :ELLIPSIS-NP T))

Now this definition is expanded. In order to describe a process as :NP-elliptic all descriptions of the subprocesses are generated as :NP's and these descriptions are then aggregated by conjunction reduction (if applicable).

Step 3:
(GOV-CONSTITUENT
   (AGGREGATION COREDAGGOBJ
      (NOUN-PHRASE :NOUN (MAKENOUN "anheben")
         :ATTRIBUTE (NOUN-PHRASE
         :NOUN "Bauchdecke"
         :PRONOUN:DEFINITE :CASE 2)
         :SUPPLEMENT (NOUN-PHRASE
            :NOUN (NOUN-PHRASE :NOUN "Backhausklemme"
               :NUMERAL "zwei")
            :PREPOSITION "mit"))))
         (NOUN-PHRASE
            :NOUN (MAKENOUN "einführen")
            :ATTRIBUTE (NOUN-PHRASE :NOUN "Kameratrokar"
            :PRONOUN :DEFINITE :CASE 2))))
         (NOUN-PHRASE :NOUN (MAKENOUN "einbringen")
            :ATTRIBUTE             (NOUN-PHRASE :NOUN "Optik"
             :PRONOUN :DEFINITE :CASE 2)))))
            :ELLIPSIS-NP T))
The respective definitions for the subprocesses are now expanded, basically giving noun phrase expressions.