Echo questions in Brazilian Portuguese^*

Perguntas-eco no português brasileiro

ABSTRACT

Brazilian Portuguese (BP) can have the wh-element in-situ with two types of sentence intonation: (a) the rising intonation of a yes/no question, in which case it is interpreted as an echo question, and (b) the falling intonation, similar to that of a declarative sentence, in which case it is interpreted as an ordinary question. Kato (2013) analyzed the falling intonation type as a fake wh-in-situ, with a short movement of the wh-element to a lower focus position, inspired by Miyagawa’s (2001) proposal for Japanese whereas the rising intonation type was analyzed in accordance with Kayne’s (1994) proposal, with the whole TP containing the wh-element moving to Spec of C. In this article we maintain the analysis of the wh-in-situ with falling intonation as a fake in-situ but analyze the echo question as a short yes/no indirect question. The languages used to support this analysis of BP are English, French, and Japanese.

Key-words:
Wh-in-situ; Echo question; Focus positions; Intonation; Brazilian Portuguese

RESUMO

O Português Brasileiro (PB) pode ter o elemento-Q in-situ com dois tipos de entoação na sentença: (a) com a entoação ascendente de uma sentença interrogativa direta sim/não, caso em que ela é interpretada como pergunta-eco e (b) com a entoação descendente de uma sentença declarativa, caso em que ela é interpretada como uma pergunta-Q ordinária. Para Kato (2013) esta última é um falso in-situ, com o elemento-Q movendo-se para uma posição baixa de Foco (cf. Miyagawa, 2001), enquanto a que se realiza com curva ascendente é analisada segundo a proposta de Kayne (1994), para quem o TP inteiro contendo o elemento-Q se move para Spec de C. No presente artigo, mantemos a análise da sentença com Q-in-situ, com curva descendente, mas reanalisaremos a sentença com interpretação de pergunta-eco como uma pergunta indireta sim/não. As línguas usadas para embasar esta análise do PB são o inglês, o francês e o japonês.

Palavras-chave:
Q-in-situ; Pergunta-eco; Posições de Foco; Entoação; Português Brasileiro

1. Introduction

English is known to have two types of wh-questions: one with the wh-element in-situ, interpreted as an echo question, and the other with the wh-element dislocated, a real question¹:

(1) a. You saw who? (echo question) b. Who did you see? (ordinary question)

(1)	a. You saw who?	(echo question)
	b. Who did you see?	(ordinary question)

Japanese, on the other hand, only presents in-situ questions, whether as an echo or an ordinary question, but the main difference lies in the overt complementizers: -tte for echo questions and -ka/-no² for ordinary questions.

(2) a. Kimi-wa dare-o mitta-tte? (echo question) you-top who-accus saw-tte b. Kimi-wa dare-o mitta-no? (ordinary question) you-top who-accus saw-no

(2)	a.	Kimi-wa dare-o	mitta-tte? (echo question)
		you-top who-accus saw-tte
	b.	Kimi-wa dare-o	mitta-no? (ordinary question)
		you-top who-accus	saw-no

Brazilian Portuguese (BP) has rising intonation (⇑) in echo in-situ questions and falling intonation (⇓) in ordinary in-situ questions (Kato, 2013a).³

(3) a. Você viu quem? ⇑(echo question) (rising intonation) you saw who b. Você viu quem? ⇓(ordinary question) (falling intonation) Lit.: Who did you see?

(3)	a.	Você viu quem? ⇑(echo question) (rising intonation)
		you saw who
	b.	Você viu quem? ⇓(ordinary question) (falling intonation)
		Lit.: Who did you see?

BP can also have cleft in-situ questions as ordinary questions (Kato, 2013a). The copula and complementizer can be erased in (4b) and (4c) (Kato, 2014):

(4) a. Foi quem que chegou? (ordinary question) was who that arrived b. (É) quem que tá tocando? (ordinary question) is who that is playing c. Quem (que) tá tocando? (ordinary question)

(4)	a.	Foi quem que chegou?			(ordinary question)
		was who that arrived
	b.	(É) quem que tá tocando?			(ordinary question)
		is who	that	is playing
	c.	Quem (que) tá tocando?			(ordinary question)

The questions to be answered are as follows:

1. What is the source of echo questions-namely, why doesn’t the wh-word move, and what determines its specific intonation?

2. What accounts for the possibility of BP to behave like Chinese or Japanese, with respect to the wh-in-situ option? Are the in-situ normal questions really comparable to the Chinese and Japanese constructions?

3. Does the wh- sit in the same position in both the echo and the regular questions in BP?

We hypothesize that the wh- in (3a) and (3b) can have different sources:

1. the echo question is an elliptical indirect speech, which accounts for the rising intonation, and the wh-constituent does not move because Comp is occupied with an embedding complementizer without the wh-feature, which attracts the wh-words; and

2. regular questions are direct questions, and I will present two alternatives for discussion: (b1), which is a slight modification of Hornstein, Nunes, and Grohmann’s (2005) analysis, and (b2), in which I claim that the wh-constituent is only apparently in-situ, undergoing a short movement to a lower FocusP position (Kato, 2013a, 2013b).

2. The Wh-parameter

2.1. [+Wh-movement] languages versus [- wh-movement languages]

Huang (1982) proposed the wh-parameter, according to which languages can have either overt wh-movement, like English, or covert wh-movement, like Japanese.

For Miyagawa (2001), the wh-phrase in English is associated with both Q-features and wh-features. Consequently, the entire wh-phrase has to move to Spec of CP to satisfy the EPP feature on C. In Japanese, the two features are distributed between two morphologically independent items: ka in C and the wh-word in T. Thus, C does not project its Specifier, as head movement can satisfy the EPP feature of C. Thus, according to Miyagawa’s view, we may have languages with [long wh-movement, to C] versus [short wh-movement, to T]. We will be assuming the latter view.

2.2. Optional [wh-movement] languages

Some languages, like French, can have both the dislocated and the in-situ types of ordinary questions and are known as optional wh-movement languages (cf Cheng & Roorick, 2000; Kato, 2013b).⁴ French is an example of such a language and exhibits the following characteristics:

1. wh-questions can have wh-fronted (with or without est-ce que), as in (5a);

2. ordinary questions can have wh-in-situ, with no visible element in Comp (5b);

3. echo questions must have wh-in-situ, with no visible element in Comp, as in (5b); and/or

4. yes/no questions have no visible element in Comp and can start with est-ce que, as in (5c).

We can add that all these examples have rising intonation.

(5) a. Qui (est-ce que) Marie a aimé⇑(ordinary or inverse-cleft wh-question) who is it that Mary has loved b. Marie a aimé qui ?⇑ (ordinary or echo question) Mary has loved who c. (Est-ce que) Marie a aimé ce garçon?⇑ (yes/no question) is it that Mary has loved this guy

(5)	a.	Qui (est-ce que) Marie a aimé⇑(ordinary or inverse-cleft wh-question)
		who is it		that Mary has loved
	b.	Marie a aimé qui ?⇑		(ordinary or echo question)
		Mary has loved who
	c.	(Est-ce que) Marie a aimé ce garçon?⇑		(yes/no question)
		is it	that Mary has loved this guy

French has restrictions for wh-in-situ constructions, as it cannot have wh-in-situ in either complement clauses, as in (6), or in islands.

(6) * Marie pense que Jean a Acheté quoi? (wh-in-situ inside complement M thinks that J has bought what clauses)

(6)	*	Marie pense que Jean a			Acheté		quoi? (wh-in-situ inside complement
		M	thinks	that	J	has bought	what clauses)

Moreover, French has rising intonation (⇑) in both ordinary and echo questions:

(7) a. Jean a acheté une voiture ?⇑ (yes/no question) John has bought a car ‘Has John bought a car?’ b. Jean a acheté quoi ?⇑ (echo or non-echo question) John has bought what

(7)	a. Jean a acheté une voiture ?⇑	(yes/no question)
	John has bought a car
	‘Has John bought a car?’
	b. Jean a acheté quoi ?⇑	(echo or non-echo question)
	John has bought what

According to Cheng and Roorick (2000), both French yes/no questions and wh-in-situ questions share the same rising intonation, due to the same Q morpheme. The presence of this Q morpheme bans the movement of the wh-word. Q can appear optionally in the numeration. If it is not in the numeration, wh-movement occurs.

3. Brazilian Portuguese (BP)

3.1. BP: An optional [wh-movement] language?

The examples in (3a) and (4b) above suggest that BP is an optional wh-in-situ language. But notice that only the echo question has the same intonation as a regular yes/no question and that the genuine question has a falling intonation (⇓). Let us compare the French example above with the BP examples below:

(8) a. Quem (que) a Maria amou?⇑ (ordinary question or cleft question) b. A Maria amou quem? ⇑ (echo question) c. A Maria amou quem? ⇓ (ordinary wh-question) d. A Maria amou este moço? ⇑ (ordinary yes/no question)

(8)	a. Quem (que) a Maria amou?⇑	(ordinary question or cleft question)
	b. A Maria amou quem? ⇑	(echo question)
	c. A Maria amou quem? ⇓	(ordinary wh-question)
	d. A Maria amou este moço? ⇑	(ordinary yes/no question)

Hornstein, Nunes and Grohmann’s (HN&G) (2005) analysis precludes optionality as the variants would be accounted for in terms of lexical choice. According to the authors, there are three Comps: one lexical que, with [+wh-feature], and two null Comps, one with [+wh-feature] and one without-namely, the in-situ one.

(9) a. [CP Quem +wh [que+wh [você viu t ] who that you saw b. [CPQuem+wh[∅+wh [ você viu t ] c. [CP ∅-wh [você viu quem]

(9)	a. [CP Quem +wh	[que+wh	[você viu	t	]
	who	that	you saw
	b. [CPQuem+wh[∅+wh [ você viu			t	]
	c. [CP ∅-wh [você viu quem]

The problem in this analysis lies in the fact that the choices are based on two phonologically identical forms with opposite values.⁵

There are, moreover, further differences between French and BP. First, BP allows wh-in-situ in embedded clauses of verbs that do not select questions as complements (10b) and only disallows wh-in-situ when the verb selects a question (10c).

(10) a. João comprou o quê?⇓ (falling intonation) John bought what b. Maria pensa que o João comprou o quê?⇓ (falling intonation) Mary thinks that the John bought what c. * A Maria perguntou se o João encontrou quem? ⇓ (falling intonation) Mary asked whether John met who

(10)	a.	João comprou o quê?⇓				(falling intonation)
		John bought what
	b.	Maria pensa que o João comprou o quê?⇓				(falling intonation)
		Mary thinks that the John bought	what
	c.	* A Maria perguntou se o João encontrou quem? ⇓				(falling intonation)
		Mary asked	whether John met	who

Second, BP allows wh-in-situ in islands:

(11) Maria pensa que o João conheceu o homem que comprou o que? (falling intonation)

(11)

Maria pensa que o João conheceu o homem que comprou o que?

(falling intonation)

To maintain HN&G’s analysis, without postulating the ∅_+wh, Kato (2013) proposed that the structure (9b) would result from a stylistic erasure of que _+wh in PF. The speaker has only two lexical choices: que _+wh and ∅_-wh .

Kato (2013a, 2013b) only analyzed the echo question as a real in-situ case, the ordinary question being a fake-in-situ, with the wh-element undergoing a short movement, like in Miyagawa’s (2001) analysis of Japanese. However, instead of moving the wh-element to T, Kato proposed that it moves to a low Focus position, in the periphery of vP. Such a position has been proposed by Belletti (2004) as an extension of Rizzi’s (1997) cartographic view for languages that can have Topic and Focus in a lower position than in the sentence periphery.

For Kato and Raposo (1996), the Focus head is a syncretic head that checks both the Focus and wh-elements.

(12) a. [ForceP... [TopP... [ FocP ... [TopP... [ Fin P... [TP ..... (Rizzi, 1997: sentence periphery) b. [CP [TP...... [TopP... [ FocP... [TopP ... [vP [VP ]]]]] (Belletti, 2004: vP periphery)

(12)	a. [ForceP... [TopP... [ FocP ... [TopP... [ Fin P... [TP	.....	(Rizzi, 1997: sentence periphery)
	b. [CP [TP...... [TopP... [ FocP... [TopP ... [vP	[VP ]]]]]	(Belletti, 2004: vP periphery)

Kato (2013a, 2013b) proposed that the CP is not the only position to check wh-features. She suggested that wh-checking, in wh-in-situ constructions in BP, do not take place at the left sentential periphery, but in a lower projection, above vP, which we assume to be FocusP (Belletti, 1998), an A’-position. Languages can be of three types:

a. only peripheral FocusP/CP (English) b. only clause-internal FocusP (Japanese ) c. both peripheral and clause-internal FocusP (BP)

a.	only peripheral FocusP/CP	(English)
b.	only clause-internal FocusP	(Japanese )
c.	both peripheral and clause-internal FocusP	(BP)

Kato (2013a, 2013b) concluded that there are no real wh-in-situ languages, but that there are languages with either long or short movement. Thus, BP would have both long and short wh-movement.

3.2. BP: A language with long wh-movement?

Until now we have been assuming that, besides a short wh-movement, BP also has a long wh-movement. Historically BP was a V2 type of language (cf. Kato & Ribeiro, 2009), with the wh-element moving to the left periphery of the sentence. After the loss of the V2 grammar in the classic period, BP started having the inverse cleft type of sentence, with the copula occupying the second position (cf. Lopes Rossi, 1996⁶), interpreted in Kato (2018) as a process of grammaticalization.

(13) a. Como veestes vós a aqueste ermo? (14th c.) how came you to this deserted place ‘How did you get to this deserted place?’ b. E quando é que são relativos? (17th c.) and when is that are relatives ‘And when is it that they are relatives?’

(13)	a.	Como veestes vós a aqueste ermo?				(14th c.)
		how	came		you to this deserted place
		‘How did you get to this deserted place?’
	b.	E quando é que são relativos?				(17th c.)
		and when is that are relatives
		‘And when is it that they are relatives?’

In the 19^th century BP started to show the fake wh-in-situ construction⁷ as a genuine question. This was followed by the appearance of canonical cleft questions, where the copula no longer appears in second position, but in first position This, in turn, was followed by the grammaticalization of the copula, which lost the consecutio temporum, becoming invariable.

(14) a. Você foi onde esta manhã? You went where this morning ‘Where did you go this morning?’ b. É onde que você foi esta manhã? Is where that you went this morning ‘Where did you go this morning?’

(14)	a.	Você foi	onde	esta manhã?
		You went	where	this morning
		‘Where did you go this morning?’
	b.	É onde	que você foi	esta manhã?
		Is where that you went		this morning
		‘Where did you go this morning?’

The canonical cleft construction follows the short wh-movement of the fake in-situ construction as the wh-element has only a short wh-movement:

(15) [TP É [FocusP onde [vP é [ FiniteP que [ TP as crianças dormem [ vPas crianças dormem [VPdormem onde ]]]]]]]

(15)	[TP É [FocusP onde [vP é [ FiniteP que [ TP as crianças dormem [ vPas crianças dormem [VPdormem onde ]]]]]]]

BP easily erases the copula in first position, even in declarative clauses, as shown in Kato (2007):

(16) a. (É) um gênio o seu filho. Is a genius the your son. ‘He is a genius, your son.’ b. (É) quem que tá tocando violão? is who that is plying guitar ‘Who is playing the guitar?’

(16)	a.	(É) um gênio o seu filho.
		Is a genius the your son.
		‘He is a genius, your son.’
	b.	(É) quem que tá tocando violão?
		is who	that is plying	guitar
		‘Who is playing the guitar?’

As canonical cleft questions appear in the language, the copula is easily erased, yielding an apparent long wh-movement with a complementizer que, in accordance with what both HN&G (2005) and Mioto (2001) proposed.

(17) a. É de que que ele está rindo? is of what that he is laughing ‘What is he laughing at?’ b. De que que ele está rindo?

(17)	a.	É de que que ele está rindo?
		is of what that he is laughing
		‘What is he laughing at?’
	b.	De que que ele está rindo?

As the inverse cleft became a canonical cleft, the wh-Focus was no longer in the periphery of the complement clause of the copula, but instead in the low vP of the copula.

Kato (2018) also proposed that the construction in (17b) can have the suppression of the complementizer que, a stylistic rule that started as a rule of haplology (avoidance of identical syllables)..

(18) a. De que que ele está rindo? > b. De que( ) ele está rindo.

(18)

a.

De que que ele está rindo?

> b. De que( ) ele está rindo.

Kato proposed that the erasure of the complementizer extended, by analogy, to other cases where the haplology condition does not apply. The final pattern in (17b), wh-SV, has been studied by Kato and Duarte(2002) as being derived directly from the V2 construction whVS, as BP changed from a [+NS] (+ Null Subject) language to a [-NS] language. As the NS parameter included VS inversion, it would be natural for the wh-SV order to be a consequence of the change in the NS parameter. The same thesis was advocated by Ordoñez and Olarrea (2006) for Caribbean Spanish, which also lost its V2 pattern in wh-questions. However, Kato (2013a, 2013b) used diachronic arguments to show that, before changing into the wh-SV order, BP had already acquired the fake wh-in-situ order and the canonic cleft pattern of Focus constructions.

Figure 1
Types of BP wh-questions through time (adapted from Kato, 2018)

Kato (2018) referenced Figure 2 to show the structural pattern that divides Modern BP from ClassicP and EP: The former has only short wh-movement whereas the latter has only long wh-movement. The structural difference is clearer before PF-namely, before copula and complementizer erasure.

Figure 2
From Old and Classic Portuguese to BP (adapted from Kato, 2018)

The conclusion is that, as far as the wh-parameter is concerned, BP is closer to Japanese and Chinese, which are both wh-in-situ languages, whereas EP pairs up with English, a wh-movement language.

4. Echo questions in Universal Grammar (UG)

At the beginning of this chapter, we pointed out that BP behaves exactly like French and English with regard to echo questions, also known as wh-in-situ constructions in the literature. These constructions present wh-in-situ with rising intonation.

(19) a. A Maria comprou o que? ⇑ b. Marie a acheté quoi? ⇑ c. Mary has bought what? ⇑

(19)	a.	A Maria comprou o que? ⇑
	b.	Marie a acheté quoi? ⇑
	c.	Mary has bought what? ⇑

In Kato (2013a, 2013b) the derivation proposed for these sentences was based on Kayne’s (1994) antisymmetry perspective. He did not differentiate between echo questions and ordinary short-movement questions. According to the author:

1. the wh-element remains in-situ; and

2. the whole TP moves to Spec, CP in such sentences.

(20) [CP [TP A Maria comprou o que ]i [ C∅ [TP ti ]]]

(20)	[CP [TP A Maria comprou o que ]i [ C∅ [TP ti ]]]

Recall that Japanese also has rising intonation in echo questions, the difference being the addition of the complementizer -tte at the end of the sentence.

(21) Mary-wa nani-o kata-tte? ⇑ Mary-top what-accus bought Comp (22) [CP [TP Mary-wa nani-o kata]i [-tte [ TP ti ]]]

(21)	Mary-wa nani-o	kata-tte? ⇑
	Mary-top what-accus bought	Comp
(22)	[CP [TP Mary-wa nani-o kata]i [-tte [ TP ti ]]]

We should also consider that, in all the languages that we have been analyzing, the echo question is identical in intonation to a yes/no question:

(23) a. A Maria comprou um IPhone? ⇑ b. Mary-wa IPHONE-o katta-no? ⇑

(23)	a.	A Maria comprou um IPhone? ⇑
	b.	Mary-wa IPHONE-o katta-no? ⇑

Although at first sight we could use Kayne’s proposal for the BP questions, in this paper we will consider a different perspective more along the lines of what happens in Japanese. We have seen the contrast between ordinary and echo questions in sentences (2a) and (2b), repeated below as (24a) and (24b). Recall that ordinary questions in Japanese are of the fake in-situ type in Miyagawa’s ( 2001) and Kato’s (2013) view, with the wh-element moving to a sentence internal Focus position.⁸ The C of the wh- clause in Japanese is overt (ka/no), as is the complementizer of echo questions (-tte); both are in the head final position.

(24) a. [Kimi-wa dare-o mitta]-tte? ⇑ (echo question) you-top who-accus saw-tte b. [Kimi-wa dare-o mitta] -no? ⇑ (ordinary question) John -top who-accus saw-ka

(24)	a.	[Kimi-wa dare-o	mitta]-tte? ⇑	(echo question)
		you-top who-accus saw-tte
	b.	[Kimi-wa dare-o	mitta] -no? ⇑	(ordinary question)
		John -top who-accus	saw-ka

When such questions appear embedded in a speech act type of a main clause, the complementizers appear overtly: -tte for the embedded echo question and -to for the ordinary wh-question. When only the embedded part appears overtly, the echo question retains the complementizer -tte of the embedded clause, and the ordinary question retains the complementizer of the main clause -no.

(25) a. [ [Kimi-wa [ dare-o mitta]-tte] yutta-no]? (indirect question) you-top who-accus saw-tte said-no ‘ You said that you saw whom?’ b. [ [Kimi-wa dare-o mitta]-tte]? b. [ [Kimi-wa [ dare-o mitta-to] yuta-no]]? (ordinary question) you -top who-accus saw =to said-no ‘John is saying that he saw whom?’

(25)	a.	[ [Kimi-wa	[ dare-o	mitta]-tte] yutta-no]?		(indirect question)
		you-top	who-accus saw-tte			said-no
		‘ You said that you saw whom?’
	b.	[ [Kimi-wa dare-o mitta]-tte]?
	b.	[ [Kimi-wa	[ dare-o	mitta-to] yuta-no]]?		(ordinary question)
		you -top	who-accus	saw	=to said-no
		‘John is saying that he saw whom?’

In BP the complementizer in the main clause is ∅₁ for echo questions and ∅₂ for ordinary questions. In BP the difference at spell-out is not morphological, but prosodic: The ordinary question appears with a falling intonation after PF whereas the echo question appears with a rising intonation.

(26) a. [CP∅1 [Você disse [CP que [TPvocê viu quem ]]]]→⇑ b. [CP∅2 [Você disse[CP que [TPvocê viu[FP [ vP quem [viu quem]]]]]]]→ ⇓

(26)	a.	[CP∅1 [Você disse [CP que [TPvocê viu quem ]]]]→⇑
	b.	[CP∅2 [Você disse[CP que [TPvocê viu[FP [ vP quem [viu quem]]]]]]]→ ⇓

Our last proposal is that the apparent simple echo question is actually a complex structure with the echo question embedded in a speech act main clause occupying the projection ForceP in the cartographic model (Rizzi, 1997).⁹ The whole derivation of (26a) is as in (27a1-a3), with the following steps:

1. the lower TP moves to Spec,ForceP

2. the Remnant TP₁ undergoes ellipsis; and

3. the ∅₁ complementizer assigns rising intonation to the resulting structure at PF.

(27) a.1. [ForceP [Force∅ [TP1 Você disse [CP que [TP2 você viu quem ]I ]]]] → a.2. [ForceP[TP2 você viu quem] [C ∅1 [TP1Você disse [CP que [ TP2 ] a.3. PF: Você viu quem? ⇑

(27)	a.1.	[ForceP [Force∅ [TP1 Você disse [CP que [TP2 você viu quem ]I ]]]] →
	a.2.	[ForceP[TP2 você viu quem] [C ∅1 [TP1Você disse [CP que [ TP2 ]
	a.3.	PF: Você viu quem? ⇑

The echo question in Japanese shows how the complementizer -tte is part of the initial structure and remains after spell-out. Recall that, being head-final, Japanese has its complementizers to the right.

(28) a. [ [Kimi-wa [kimi-ga dare-o mitta-tte] yutta]-no]? (Indirect question) you-top you-nom who-accus saw-tte said-no (29) a. [ [Kimi-ga dare-o mitta]-tte] (Echo question}

(28)	a.	[ [Kimi-wa	[kimi-ga dare-o mitta-tte]	yutta]-no]?	(Indirect question)
		you-top you-nom who-accus saw-tte		said-no
(29)	a.	[ [Kimi-ga dare-o	mitta]-tte]		(Echo question}

English, French, and Portuguese do not exhibit a complementizer like Japanese, which has the complementizer -tte that also appears in complex indirect questions. This led to the hypothesis that echo questions are actually reduced indirect questions that also have rising intonation.

5. Conclusions

This study has shown how far we have come from the classic wh-parameter when we examine languages as different as BP and Japanese, with the addition of a more well-behaved language like English.

We first tried to see whether BP was a partial type of [-wh-movement] language, like French. Failing to find any similarity, we decided to compare it with a distant language, like Japanese, with which we had previously compared it in terms of a fake in-situ language, regarding ordinary wh-questions with short wh-movement (Kato, 2013). Having been successful in that enterprise, we decided to compare the two languages for echo questions.

What we found is that the echo type was much more universal, with the languages sharing the same rising intonation. Using the overt complementizer -tte in Japanese, we proposed that the echo type can be analyzed as a covert embedded clause that is moved to the Spec of a higher projection, the ForceP, after which the TP1 undergoes remnant ellipsis.

References

• BELLETTI, A. 2004. Aspects of the low IP area. In: L. Rizzi (Ed.). The structure of IP and CP: The cartograph of syntactic structures. New York: Oxford University Press. p. 16-51.
• BOŠKOVIČ, Ž. 1998. LF movement and the minimalist program. In: P. N. Tmanji; K. Kusumoto (Eds.). Proceedings of NELS 28. Amherst, MA: GLSA. p. 43-57.
• CHENG, L. S.; ROORYCK, J. 2000. Licensing wh-in-situ. Syntax, 3(1), 1-19.
• DUARTE, M. E. L.; KATO, M. A. 2002. A diachronic analysis of Brazilian Portuguese wh-questions. In Santa Barbara Portuguese studies. Vol. VI. Santa Barbara, CA: University of California at Santa Barbara, Center for Portuguese Studies. p. 326-339.
• HORNSTEIN, N.; NUNES, J.; GROHMANN, K. K. 2005. Understanding minimalism. Cambridge, UK: Cambridge University Press.
• HUANG, C.-T. J. 1982. Logical relations in Chinese and the theory of grammar (Unpublished Ph.D. dissertation). MIT, Cambridge, MA.
• KATO, M. A. 2007. Free small clauses in BP. DELTA, 23, p. 85-111.
• KATO, M. A. 2013a. Wh-movement in the history of Brazilian Portuguese. STIL, 5, p. 59-76. Retrieved from http://www.ciscl.unisi.it/doc/doc_pub/
  » http://www.ciscl.unisi.it/doc/doc_pub/
• KATO, M. A. 2013b. Deriving wh-in-situ through movement. In: V. Camacho-Taboada; A. Gimenez-Fernancez; J. Martin-Gonzales; M. Reyes-Tejedor (Eds.). Information structure and agreement. Amsterdam/Philadelphia: John Benjamins. p. 175-191.
• KATO, M. A. 2014.The role of the copula in the diachronic development of focus construction. In: M.-H. Côté; E. Mathiew (Eds.). Variation within and across Romance languages. Amsterdam/Philadelphia: John Benjamins. p. 294-314.
• KATO, M. A. 2018. Morphological doublets in Brazilian Portuguese wh-constructions. In: L. Repetti; F. Ordoñez (Eds.). Romance languages and linguistic theory 14. Amsterdam/Philadelphia. John Benjamins. p. 135-152.
• KATO, M. A.; RAPOSO, E. 1996. European and Brazilian word order: questions, focus and topic constructions. In: C. Parodi; A. C. Quicoli; M. Saltarelli; M. L. Zubizarreta (Eds.). Aspects of Romance linguistics. Washington, DC: Georgetown University Press. p. 267-277.
• KATO, M. A.; RIBEIRO, I. 2009. Cleft sentences from old Portuguese to modern Brazilian Portuguese. In: A. Dufter; D. Jacob (Eds.). Focus and background in Romance languages. Amsterdam/Philadelphia: John Benjamins. p. 123-154.
• KAYNE, R. 1994. The antisymmetry of syntax. Cambridge, MA: MIT Press.
• MIYAGAWA, S. 2001. The EPP, scrambling, and wh-in situ. In: M. Kenstowicz (Ed.). A life in language Ken Hale. Cambridge, MA: MIT Press. p. 293-338.
• NOONAN, M. 1989. Operator licensing and the case of French interrogatives. Proceedings of the 8th West Coast Conference on Formal Linguistics. p. 315-330.
• ORDOÑEZ, F.; OLARREA, A. 2006. Microvariation in Caribbean/non Caribbean Spanish. Probus, 18(3), p. 59-97.
• RIBEIRO, I. 1995. Evidence for a verb-second phase in old Portuguese. In: A. Battye; I. Roberts (Eds.). Clause structure and language change. New York: Oxford. p. 110-139.
• RIZZI, L. 1997. The fine structure of the left periphery. In: L. Haegeman (Ed.). Elements of grammar. Dordrecht: Kluwer. p. 281-337.

• 1
  Wh-in-situ is also present in multiple wh-questions. (i) Who bought what?
• 2
  .-no is the polite version of -ka.
• 3
  We consider only the final contrastive intonation at the end, and not other possible intonational differences elsewhere in the sentence.
• 4
  See also Boskovič (1998), who analyzes French as having LF insertion of C^o with a strong wh-feature.
• 5
  See a similar view in Mioto (2001), for whom there is also the insertion of ?_+wh before spell-out.
• 6
  Old Portuguese (OP) also exhibited the pseudo-cleft type of wh-question, like in Spanish, but due to the scarce occurrence of this construction, we will not include it in our analysis.
• 7
  Lopes Rossi (1996) did not consider the in-situ cases as fake in-situ.
• 8
  Recall that, for Miyagawa (2001), the wh-element moved to a position adjacent to T. We proposed that in BP it moved to a Focus position adjacent to vP.
• 9
  The cartographic model, which started with Rizzi (1997), has other projections like TopP and FocP that can be read in Kato (2013b, 2018).
• *
  I would like to thank Marcello Marcelino for his usual help in revising my English.

Publication Dates

• Publication in this collection
  06 May 2019
• Date of issue
  2019

History

• Received
  20 Aug 2018
• Accepted
  14 Oct 2018