Natural languages form questions in at least two different ways: question words appearing at the beginning of the sentence, see the contrast in English examples (1) and (2), vs question words appearing in situ, causing no change in the word-order, see Chinese examples (3) and (4). (1) Who did John see? (2) John saw Mary. (3) John kanjia-le shei Who did John see? (4) John kanijia-le Mary John saw Mary. These two cases have been investigated by syntacticians since the 1980s and have been argued to have two different sets of properties. Under the view that all question words must move to the beginning of the sentence to get their meaning, it was proposed that question words move in cases like (3) too, however in such cases this movement happens in a way that it affects only meaning and not sound whereas in cases like (1), the movement of the question word affects both sound and meaning. In this model, these two cases also differ in terms of the conditions that constraint them. It was believed that movement in cases like (1) is more constrained than movement in cases like (3), as the former is constrained by conditions imposed by both the sound and meaning component of the grammar, whereas the latter is constrained only by the conditions imposed by the meaning component. This study challenges this traditional view while investigating question formation in Hindi, a language where movement of question words seems to affect only meaning, see (5) and (6). However against expectations, this movement obeys the same constraints that are attributed to movement that affects both sound and meaning. This poses a challenge to the existing model. (5) John-ne kis-ko dekha Who did John see? (6) Johnne Mary-ko dekha John saw Mary. This research takes this finding to propose that movement of the question words essentially happens in the same way in all languages and obeys the same constraints, even though it is silent in some cases and gets detected in others.