by Hiroaki Teraoka
This is a reproduction of my term paper for an English linguistic course. The original paper is presented in 2022.
1 Present Day Englis
According to Chomsky (1995), we need an operation merge to build phrases and sentences. We merge two constituents (i.e., words or phrases) to form larger phrases. For example, a present-day English sentence he used this pen can be analyzed as follows:
(1) He used this pen.
(2) The internal structure of (1)
We merge a noun phrase (NP) pen with a determiner to form a determiner phrase (DP) this pen. The determiner determines the grammatical feature of the overall phrase. Thus, the determiner is the head of the DP. The NP is the complement of the D. We merge the resulting DP with a verb use to form a VP use this pen. This time, the V use is the head of the overall phrase. The DP is the complement. Then we merge this VP with the tense (T) suffix -ed to form a T-bar -ed use this pen. We merge this T-bar with the specifier (he) to form the complete TP he -ed use this pen. We merge this TP with a null complementizer to form a CP. When this CP is pronounced, the verbal suffix -ed is lowered to the V head position by affix hopping (Chomsky 1957).
When you make a yes-no question sentence out of (1), you get (3).
(3) Did he use this pen?
The reasonable question arises here: why we use do-support. To answer this question, we need to analyze the internal structure of (3).
(4) The internal structure of (3).
According to Radford (2016), we move constituents in T to C to make a yes-no question sentence. In this case, the T has verbal suffix -ed. We move this verbal suffix to the immediately higher head above T, which is the C. Chomsky (1981) claims that a head cannot jump over another head. Thus, the constituent in the T can move to the C. The verbal suffix in the C has no verb host. Thus, we insert a semantically null auxiliary verb do to save the verbal suffix in the C. Linguists call this operation do-support (Radford 2016).
We use do-support in negative clauses.
(5) He did not use this pen.
The reasonable question is why we use do-support in negative constructions in present-day English. To answer this question, we need to analyze the internal structure of (5).
(6) The internal structure of (5)
Radford (2016) introduces negative phrases (NegPs). In present-day English, a negative phrase has null constituent as its head and not as its specifier. The null head of the NegP takes a VP as its complement. Thus, we merge the VP use this pen with the null negative head to form Neg-bar. We merge this neg-bar with the specifier (not) to form a full negative phrase. Then, we merge this NegP with the T suffix -ed to form a T-bar. We merge this resulting T-bar with the specifier (he) to form a full TP. When we pronounce this clause, we try to lower the tense suffix -ed to the V head position by affix hopping. However, we cannot. Chomsky (1981) claims that a head can only move to an adjacent head. Since the negative phrase has its own head, the tense suffix in the T need to jump over the negative head to reach the V head. Such movement is banned. Thus, the verbal suffix is stranded in the T head position. To save this tense suffix, we insert a semantically null auxiliary verb do when we pronounce the clause.
2 Middle English
In this section, I cite several examples from Chaucer to see how speakers of Middle English made question sentences and negative sentences.
(7) Why sholde men thanne speke of it vileynye?
‘Why should men then speak evil of it?’
(Chaucer. Canterbury Tales. The Prologe of the Wyves Tale of Bathe. Line 34. From Harvard’s Geoffrey Chaucer Website. URL: https://chaucer.fas.harvard.edu/pages/wife-baths-prologue-and-tale-0)
In (7), the auxiliary verb should precedes the subject (men) in the specifier of the TP. Thus, The auxiliary originated in the T moved to the C position, which precedes the spec-TP. The wh-question phrase why precedes the C. Thus, why is placed in the spec-CP. When a clause has an auxiliary verb, the word order of a wh-question is the same as that of present-day English.
However, things get different when a clause lacks auxiliaries.
(8) What rowne ye with oure mayde? Benedicite!
‘What do you whisper with our maid? Bless me!’
(ibid. Line 241.)
Although I looked the word rowne in Oxford English Dictionary, I did not find it. Judging from the translation, rowne is a verb with its suffix. If we analyze the ne as the suffix, the internal structure of the (8) is diagramed as follows:
(9) The internal structure of What rowne ye with oure mayde
The verb row merges with what to form a VP row what. We merge this VP with a PP with oure mayde to form an even larger VP row what with oure mayde. We merge the resulting VP with the tense suffix -ne in T to form a T-bar. According to Roberts (2007, 2021), Ts in Middle English were ‘strong’. What Roberts means by ‘strong Ts’ is that verbal suffixes in Ts were morphologically rich in Middle English and they attracted the movements of constituents in Vs to Ts. Thus, the verb originated in the V moves upward to the T to receive the verbal suffix -ne. Roberts (2007, 2021) call this movement of the verb V-to-T movement. We merge the resulting T-bar with its specifier (ye) to form a full TP. We merge this TP with a null C. This null C attracts the constituents in the T to the C position. This movement is the same as T-to-C movement in present-day English. According to Roberts (2016) when constituents in a head position move, they cannot leave anything. Thus, we move the verb row and its suffix -ne to the C. Row+-ne are now in the C position. We merge this C-bar with specifier (what), which has moved from inside the VP. When we pronounce the resulting CP, we get What rowne ye with oure mayde.
In Middle English, a verb originated in V moved through T to C. The verbal suffix originated in T moved with the verb to C. C had verbs and verbal suffix. Thus, in Middle English, you did not need do-support to make a question clause.
Middle English had negative concords. In a negative concord construction, a sentence has two negative words but the meaning of the overall sentence is still negative. I cite an example of a negative concord from Chaucer.
(10) I ne owe hem nat a word that it nys quit.
‘I owe them not one word that has not been avenged.’
(Chaucer. Canterbury Tales. The Wyves Tale of Bathe. Line 424. Ibid.)
Although this example has two negative words (i.e., ne and nat), the meaning of the overall clause is negative, not declarative. According to Bybee (2010), the negative word not has developed from no or na + wiht ‘someone, something’. This negative word not will be fully grammaticalized and placed in the specifier of a negative phrase. However, in this example, the negative word nat was not fully grammaticalized. This word seems to work as a negative quantifier. Thus, we can analyze nat a word that it nys quit as a quantifier phrase (QP).
(11) The internal structure of the vP part of I ne owe hem nat a word
The clause has three arguments. Thus, I need to use vP analysis. The verb owe merge with the QP to form the V-bar. We merge the resulting V-bar with its specifier hem to from a full VP. The VP is in turn merged with null v. This v attracts the constituent in V to itself. Then, we merge the resulting v-bar with its specifier. Thus far, we have made I owe hem owe nat a word… The second instance of owe is the copy and it is silenced when pronounced (Chomsky 1995). According to Radford (2009), Larson thought up this vP (pronounced as ‘small VP) analysis.
We merge the resulting vP with a negative head ne.
(12) The internal structure of the second half of I ne owe hem nat a word
The head of the negative phrase ne attracts the constituents in the v head position. Thus, owe in v moves to the negative head position. The negative phrase in Middle English seems to lack a specifier. Thus, we merge the negative phrase with the tense suffix. According to Roberts (2007, 2021), T in Middle English was strong. Thus, the T attracts the constituents from immediately below head. The negative head has ne owe. Since head movements cannot leave anything (Radford 2016), we move both ne and owe to T. Chomsky (1995) argue that Ts in present-day English have the EPP feature. EPP stands for extended phrasal projection. This means that Ts in present-day English must have specifiers. Ts in Middle English seems to have this EPP feature. Thus, the EPP feature in T attracts a suitable candidate to the specifier of the TP. The EPP feature search its domain. The domain of the T is T’s complement, namely, the negative phrase. In this case, the EPP feature in the T finds a suitable candidate in the specifier of the vP and attracts it to the specifier of the TP. In this way, we have got the correct word order for I ne owe hem nat a word.
3 Early Modern English
In this section, we will see how word order changed in Early Modern English.
(13) Can you tell if Rosalind the Dukes daughter bee banished with her Father?
(Shakespeare. As You Like It. Spoken by Oliver. Taylor, G., Jowette, J., Bourus, T. and Egan, G. eds. 2017: 1867)
(14) Will you go to them?
(Shakespeare. Romeo and Ivilet. Spoken by Nurse. Ibid. 724)
Speakers of Early Modern English (which was spoken form 1500 to 1700) made yes-no questions by T-to-C movement. In other words, they moved constituents in Ts to Cs to form yes-no question clauses. This operation has not changed from Middle English. Even present-day English speakers use the same T-to-C movement to make question root clauses.
Speakers of Early Modern English made wh-questions also by T-to-C movements and placing wh-phrases in specifiers of CPs.
(15) How will she none?
(Shakespeare. Romeo and Ivilet. Spoken by Capulet. Ibid. 734)
The word none seems to be used as a verb. However, even though I looked the word in Oxford English Dictionary, I found no instance of the word used as a verb. Nevertheless, auxiliary verb originated in the T precedes the subject in the specifier of the TP. Thus, the auxiliary moved from the T to the C position. The wh-phrase how precedes will in the C. Thus, how is in the specifier of the CP. Thus far, every example I cited from Shakespeare had an auxiliary verb. I deal with word orders of the clause which had no auxiliaries.
Although suffixes in Ts were strong in Middle English, they got weaker as time passed.
(16) does she not give us thanks? (ibid.)
(17) does she not count her blest? (ibid.)
(16) and (17) are examples from Shakespeare. These examples clearly show that Ts in Early Modern English were weak. The internal structure of (16) in diagramed in (18).
(18) The internal structure of (16).
The T was weak in Early Modern English. Thus, the verb remained in its original position, namely, the V. In Early Modern English, the negative phrase seems to have had null constituent as its head and not as its specifier. The null head of the NeegP takes the VP as its complement. Then, we merge not as the specifier of the NegP. The T head -es takes the NegP as its complement. We merge a pronoun she with the resulting T-bar to form a TP. We merge this TP with a null C. We are making a yes-no question clause. Thus, this null C attracts the constituent in T. The verb remains in the V head. Thus, the C attracts verbal suffix -es to itself. The verbal suffix must be pronounced with a host verb. Thus, the speakers of Early Modern English inserted do to save the verbal suffix.
4 Conclusion
We have seen haw word order changed from Middle English to Early Modern English. Middle English did not use do-support for question clauses or negative clauses. Roberts (2007, 2021) claims that Ts in Middle English were strong and attracted the movements of constituents in Vs. Because of this V-to-T movement, the verbal suffixes were never stranded. Thus, speakers of Middle English did not need do-support. This strong T analysis of Roberts also explains the position of a negative word ne, which preceded main verbs. Early Modern English had weak Ts. According to Roberts (2007), the strong T parameter determines whether Ts are strong or not in a language. The value of this parameter seems to have changed in Early Modern English. Shakespeare used do-supports for question clauses and negative clauses. This shows that Ts in Early Modern English were weak and they did not cause V-to-T movements.
Primary Source)
Harvard’s Geoffrey Chaucer Website. URL: https://chaucer.fas.harvard.edu/ (last accessed 5th August, 2022)
Taylor, G., Jowette, J., Bourus, T. and Egan, G. (eds.) (2017) The New Oxford Shakespeare: The Complete Works: Critical Reference Edition. 2vols. Oxford: Oxford University Press.
Secondary Source)
Bybee Joan (2010) Language Usage and Cognition. Cambridge: Cambridge University Press.
Chomsky, N (1957) Syntactic Structures. Leiden: Mouton and Co.
Chomsky, N (1981) Lectures on Government and Binding – The Pisa Lecture. Hague: Mouton de Gruyter. (formerly published by Foris Publications)
Chomsky, N. (1988) Language and Problems of Knowledge : The Managua Lectures. Cambridge: MIT Press.
Chomsky, N. (1995) The Minimalist Program. Cambridge, MA: MIT Press.
Hopper, P. J. and Traugott, E. C. (1993) Grammaticalization. Cambridge: Cambridge University Press.
Kegl, J., Senghas, A. and Coppola, M. (1999) “Creation through Contact: Sign Language Emergence and Sign Language Change in Nicaragua.” in DeGraff, M. (eds.) (1999) Language Creation and Language Change—Creolization, Diachrony, and Development. Cambridge, Massachusetts: MIT Press. 177-237.
Oxford English Dictionary. https://www.oed.com/
Radford, A. (2009) Analyzing English Sentences. Cambridge: Cambridge University Press.
Radford, A. (2016) Analyzing English Sentences. 2nd edition. Cambridge: Cambridge University Press.
Roberts, I. (2007) Diachronic Syntax. Oxford: Oxford University Press
Roberts, I. (2021) Diachronic Syntax. 2nd edition. Oxford: Oxford University Press.
