The aim of the transformational analysis is to discover how the sentence under analysis is generated. To achieve this objective we have to find the kernel structures which are the part of the deep structure of the sentence we analyse and the transformations that change the kernel structures into the surface sentence.
The analysis can proceed either from a set of kernel structures and develop them to the sentence which is analysed. The analysis can also proceed either from the sentence which is analysed to a set of kernel structures which are developed into the sentence.
We can take for illustrating both ways of analysis the sentence "Riccardo was a LITTLE surprised at their keeping the whole committee waiting to look at a strolling company of mountebacks".
A.From kernel structures to the sentence.
Since every kernel structure possesses a VP, we may assume that there are as many such structures as we find verbal forms or derivatives of verbs. The sentence contains five verbal forms: be surprised, keeping,waiting, to look, strolling. It means that the deep structure of the sentence includes at least five kernel structures. These structures are:
1. (NP1+BE+AP1)
2. (NP2+VP1+AdvP)
3. (NP4+VP2)1
4. (NP5+VP3+prpNP6)
5. (NP7+VP4)2
The symbols here mean: VP — verb phrase; NP — noun phrase; AP — adjective phrase; AdvP — adverbial phrases; the figures are used to differentiate phrases from each other.
The next step is identification of the phrases which include only ultimate constituents. We can find such phrases: NP1 = Riccardo; BE = was;; NP4 = committee; VP2 = wait; VP3 = look; VP4 = stroll. Inserting them into our kernel structures we receive:
1. (Riccardo +was+AP1)
2. (NP2+ VP1+(NP12+ AdvP)
3. (committee + wait)1
4. (NP5+ look +prpNP6)
5. (NP7+ stroll)2
At the next step we are to develop the constituets left unidentified.
(Riccardo +was+AP1)
AP1—> A+prp+NP8
NP8—>(NP2+ VP1+AdvP)
NP2—>they
VP1—>VP5+NP9
AdvP—>(NP5+ look +prpNP6)
NP5—>they
prpNP6—>(NP7+ stroll)2
NP7—>NP10+prpNP11
NP10—>Det+NP10—>a company
prp—>of
NP11—>mountebacks
NP12—>(NP4+VP2)1—>
Now we can perform substitution of the constituents of the kernel structures with their developed forms.
5. (NP7+ stroll)2—>( a company of mountebacks stroll)2
4. (NP5+ look +prpNP6)—> (NP5+ look +prp(NP7+ stroll)2)—> (NP5+ look +prp ( a company of mountebacks stroll)) —> (they + look +at ( a company of mountebacks stroll))
3. (NP4+VP2)1—>(committee + wait)
2. (NP2+VP1+AdvP)—> (NP2+ (VP1+NP2)+ AdvP)—> (NP2+ (VP1)+(NP5+ look +prpNP6))
1. (NP1+BE+AP1)—>(NP1+was+A+prp (NP2+(VP1+(NP3+VP2)+(NP2+ VP3+(NP$+VP4)—>Riccardo was /a little/ surprised at (they keep (the/whole/ committee wait) to (they look at (a company of mountebacks stroll))).
At this stage the procedures of the second type, of the type altering the kernel structures should be used. Generally the procedures are applied to the symbols representing the appropriate constituents. But this way of transforming demands lengthily explanations. To avoid these long passages of that this N is, say, "company" and this V is, say, "wait" etc, we shall apply the transformations to actual structures with real words in them.
Riccardo was /a little/ surprised at (they kept (the/whole/ committee waited) (they looked at (a company of mountebacks strolled))).
The first operation should be applied to the constituent (a company of mountebacks strolled). This constituent in the larger one functions as an N, but it is a clause at the deep structure level. It means that we have transform it into a noun. It is done in the following way:
a) we change the finite form into a Participle I, for which we substitute |ed| by |ing|: (a company of mountebacks strolled)—> (a company of mountebacks strolling);
b) to acquire the correct form of the phrase we transmutate the constituents (company of mountebacks) and (strolling): (a company of mountebacks strolling)—> (a strolling company of mountebacks)
c) to complete the operation we remove the parentheses: (they looked at (a strolling company of mountebacks))—> (they looked at a strolling company of mountebacks)
The next is the constituent (they looked at a strolling company of mountebacks), which is already partially modifies. This constituent forms an adverbial phrase, but at the deep structure level is again a clause. The task is to transform it into an equivalent of an Adverb. For that we have to substitute the finite form "looked" by a non-finite, this time it should be an Infinitive. Again substitution of the grammatical indicator is done and we receive (they to look at a strolling company of mountebacks). As yet we cannot do anything with the Personal Pronoun, but this will be solved when we complete with the larger unit of which this constituent is a part.
The next constituent to be processed is (the/whole/ committee waited). This constituent is in the position of a noun phrase in the larger structure, and we have to make it look like a noun. And again here we should modify the verbal element of the clause transforming it this time into a Gerund by substitution of |ed|: (the/whole/ committee waited)—> (the/whole/ committee waiting).
Now we can insert the results into the larger constituent and receive the phrase: (they keep (the/whole/ committee waiting) (they to look at a strolling company of mountebacks)). When we remove parentheses we receive the structure (they keep the whole committee waiting they to look at a strolling company of mountebacks), in which we find repetition of the same element "they". To make the structure correct we have to delete the second "they" and finally have (they keep the whole committee waiting to look at a strolling company of mountebacks).
Now we should see if it is necessary to modify the constituent we have just produced. If we look at the whole structure we notice that again it function is not of a clause which it is at the deep structure level, but that of a noun, and again we have to transform a clause into a noun. The procedure has already been described: we should substitute the indicators of the finite form by those of the non-finite, this time of the Gerund. At the same time we have to substitute the Nominative Case of the Personal pronoun by its Objective Case. The ultimate result is:
Riccardo was a little surprised at their keeping the whole committee waiting to look at a strolling company of mountebacks.
Some comments should be added to the description of transformations used in the example. We may notice that in our procedure we employed several times similar complexes of operations . Such stable complexes of elementary operations are often referred to by special names and used as a block. In our example these were a block of operation which is named by its result — Nominalisation (making a form functioning as an NP) and Adjectvisation (making a form functioning as an AP). Both block {or standard transformations) have several stages which are distinguished by the number of substitution operations employed and the results achieved. Nominalisation transformation in English has the stages
1. stage, generation of the Infinitive: describes—> (to) describe:
2. stage, generation of the Gerund: (to) describe—>describing:
3. stage, derivation of the Noun: describing—>description.
Adjectivisation transformation has only two stages:
1. stage, producing a Participle: describes—>describing or described;
2. stage, making Adjectives: describing—>descriptive or described—>describable.
The transformational analysis has many such stable complexes of operations. The most frequently employed are three — Nominalisation, Adjectivisation, and Passivisation (making a Passive construction out of Active)
B. From the sentence to kernel structures
If we take the same sentence for detecting its deep structure we have to start with the Immediate Constituents analysis. It results in the following