<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.2 20190208//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:ali="http://www.niso.org/schemas/ali/1.0">
  <front>
    <article-meta>
      <article-categories>
        <subj-group>
          <subject content-type="Tipo de contribuio">Ensaio teórico</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>COUNTERDIRECTIONALITY CROSS- LINGUISTICALLY</article-title>
        <subtitle>COMPARING BRAZILIAN PORTUGUESE AND DUTCH</subtitle>
      </title-group>
      <contrib-group content-type="author">
        <contrib id="person-3de55a27984fd9273addbdc012efc9ec" contrib-type="person" equal-contrib="yes" corresp="no" deceased="no">
          <name>
            <surname>ZWARTS</surname>
            <given-names>Joost</given-names>
          </name>
          <email> leedimufscar@gmail.com</email>
          <xref ref-type="aff" rid="affiliation-8dd668946216ad4a58fc60b24859a29e" />
        </contrib>
        <contrib id="person-0779208b934212dd33563df3a197b486" contrib-type="person" equal-contrib="yes" corresp="no" deceased="no">
          <name>
            <surname>BASSO</surname>
            <given-names>Renato Miguel</given-names>
          </name>
          <xref ref-type="aff" rid="affiliation-cd54d3f1df2b61c448226095bcdb162e" />
        </contrib>
      </contrib-group>
      <aff id="affiliation-8dd668946216ad4a58fc60b24859a29e">
        <institution content-type="orgname">Utrecht University</institution>
      </aff>
      <aff id="affiliation-cd54d3f1df2b61c448226095bcdb162e">
        <institution content-type="orgname">Universidade Federal de São Carlos (UFScar)</institution>
      </aff>
      <pub-date date-type="pub" iso-8601-date="01/04/2016" />
      <volume>15</volume>
      <issue>1</issue>
      <issue-title>COUNTERDIRECTIONALITY CROSS- LINGUISTICALLY: COMPARING BRAZILIAN PORTUGUESE AND DUTCH</issue-title>
      <fpage>19</fpage>
      <lpage>44</lpage>
      <page-range>19-44</page-range>
      <history>
        <date date-type="accepted" iso-8601-date="05/11/2015" />
        <date date-type="received" iso-8601-date="01/09/2015" />
      </history>
      <permissions id="permission">
        <license>
          <ali:license_ref>http://creativecommons.org/licenses/by/4.0/</ali:license_ref>
        </license>
      </permissions>
      <abstract>
        <p id="_paragraph-1">
          <italic id="italic-1">In this paper, we investigated the expression of “counterdirectionality”, contrasting Brazilian Portuguese and Dutch data. Firstly, we define counterdirectionality using English data, and we present some of its subdivisions in meaning. Secondly, we investigated the expression of counterdirectionality in Brazilian Portuguese and Dutch, comparing the two languages.</italic>
        </p>
      </abstract>
      <abstract abstract-type="executive-summary">
        <title>Resumo</title>
        <p id="paragraph-4aeff869000d1770d78141f8f76a5d94">
          <italic id="italic-d600e05b42b0c8b60d88586720a463ee">Neste artigo, investigamos a expressão da “contradirecionalidade”, contrastando dados do português brasileiro com o holandês. Num primeiro momento, definimos contradirecionalidade usando dados do inglês, e apresentamos algumas de suas subdivisões que serão então investigadas no português brasileiro e no holandês.</italic>
        </p>
      </abstract>
      <kwd-group>
        <kwd content-type="">
          <italic id="italic-0090cb889184a2b57f4a343355741347">Contradirecionalidade</italic>
        </kwd>
        <kwd content-type="">
          <italic id="italic-26e79fb4e4159a7ed12fffbf993e3f32">construções espaciais</italic>
        </kwd>
        <kwd content-type="">
          <italic id="italic-568857670305621cf92f6bcef07cbea5">preposição</italic>
        </kwd>
        <kwd content-type="">
          <italic id="italic-be67ebec06b1c1ed4d6f9bfd65138cb2">semântica</italic>
        </kwd>
        <kwd content-type="">
          <italic id="italic-581605a8a8fc373fdddcd273f0dbe9f7">metáfora</italic>
        </kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body id="body">
    <sec id="heading-f389db4b7cb1b0eacccbde541223f8e7">
      <title>Introduction</title>
      <p id="paragraph-1">In the semantics of verbs and verbal modification, there is a domain that we might characterize, following Fabricius-Hansen (2001) as <italic id="italic-6d77df6d0cff5b621ed3187361f86cb3">counterdirectional</italic>. It involves the description of processes that are in some sense opposite in direction to what is common or presupposed. We find in this domain such expressions in English as the adverbs <italic id="italic-2">back</italic>, <italic id="italic-3">backward(s)</italic>, and <italic id="italic-4">again </italic>and the prefixes <italic id="italic-5">re- </italic>and <italic id="italic-6">counter-</italic>. A detailed study of <italic id="italic-7">back </italic>is offered in Allan (1995) from a cognitive linguistic perspective. There are some typological studies of this domain (Lichtenberk 1991, Wälchli 2006, Moise-Faurie 2012); the concept BACK also plays a role in grammaticalization theory (Heine &amp;Kuteva 2002). The most intensive study is made at the interface of semantics and syntax, with a focus on the so-called repetitive and restitutive meanings of <italic id="italic-8">again </italic>and its German counterpart <italic id="italic-9">wieder </italic>(e.g., Dowty 1979, Von Stechow 1996, Beck 2005, Pedersen 2014).</p>
      <p id="paragraph-2">In this paper we want to contribute to the understanding of this domain on the basis of a comparison of Brazilian Portuguese (BP) and Dutch. Based on a rough inventory and map of counterdirectional senses that we identify for English (section 2), we want to determine what type of markers these languages use to express counterdirectionality and how these markers distribute over the map of senses (section 3). Our goal is to show how these two languages divide up the same set of meanings in sometimes different ways, exhibiting opposite diachronic developments for some of the markers (section 4). Since the concept of counterdirectionality has not been extensively discussed and defined in the literature and is still somewhat vague, another goal of this paper is to make the domain of counterdirectional meaning more precise by replacing it with a family of well-defined semantic categories that are then applied to the two languages in focus.</p>
    </sec>
    <sec id="heading-b144f83a20159a0c660ef0416292560e">
      <title>1. Counterdirectional senses in English</title>
      <p id="paragraph-fa2c31c82bbf277da32682452ef8e4d5">There is not a systematic semantic study of the different senses in the counterdirectional domain. What senses need to be distinguished and how do these senses hang together? All studies distinguish the restitutive and repetitive sense, but apart from that, there is less consensus, which makes the notion of counterdirectionalitystill somewhat vague. It is important to realize that counterdirectionality is not one single concept, but a family of closely related semantic categories. What we offer in this section is an overview of six senses from this family, inspired by the existing literature. English will be our object language in this section and we focus, more specifically, on the adverbs <italic id="italic-7e5ebfb682834a848c61cb72461c3e4d">back</italic>, <italic id="italic-2430e49dc6f10000820394450fa5d010">backward(s)</italic>, and <italic id="italic-8d7af35dc804221a4146d88db8709d4b">again </italic>when used in combination with verbs or verb phrases. We identify six senses: rearward, retrograde, returnative, responsive, restitutive, and repetitive.</p>
      <p id="paragraph-a7c5f7f99df36ef44717267f97c66dd2">The rearward sense obtains most clearly when a person moves or leans in the direction of his or her back, that is, with the back leading. This is where English has <italic id="italic-69a77ec2eca2b167773d99c01b32fd02">back </italic>or <italic id="italic-9e39d2529e5e607ec63454f8a52e7d44">backward(s) </italic>(Allan 1995).</p>
      <p id="paragraph-6">(1)</p>
      <p id="paragraph-7">a. Dunworthy stepped <italic id="italic-2b142fcac7451a0bc83b28d5ecb5bdd5">backward </italic>and crashed into a six- year-old holding a plush Santa. (Connie Willis – The Doomsday Book)</p>
      <p id="paragraph-8">b. In the hunting lodge at Dachigam, Max Ophuls reclining on carpets and cushions leaned <italic id="italic-12bcfa660abd524ee7b0796c02551913">backward</italic>, away from the Indian foreign minister, to whisper to Edgar Wood. (Salman Rushdie – Shalimar The Clown)</p>
      <p id="paragraph-9">c. Gravity dragged at her and she fell <italic id="italic-669f93dc833da6d6e8d6fe618c6ca9d6">backward </italic>on the bed, gasping for air. (Salman Rushdie – Shalimar The Clown)</p>
      <p id="paragraph-11">d. I sat <italic id="italic-2e0dce71af55dca727f9b4028de25efe">back </italic>in my chair.</p>
      <p id="paragraph-61e4b375a1fc4fb16892b5f0b6a8d1ac">.</p>
      <p id="paragraph-de3faa5edacc19e8e171ceb26548f7fb">It seems that <italic id="italic-cafb43f73d12f9a576a5df63b85ddcd7">backward(s) </italic>is the productive adverb for this sense and that <italic id="italic-eaec271aea90153ae3151d4ce435c779">back </italic>is the more lexicalized form, restricted to certain verbs.</p>
      <p id="paragraph-e0dc1553a05e7d2e57533d58f5efac24">Note that these expressions can be used to describe a situation in which a person goes to places where he or she has not been before. This makes the rearward sense different from the returnative sense that we will discuss below. In order to clarify what we should understand by rearward (and also the other senses that we’ll see shortly), consider the following figures. In these figures we have a sketch of a path, with a source and a goal, an arrow above the figure indicating direction of movement, and finally the direction which the figure is facing. For the rearward sense, we have two possibilities, as shown by figures 1 and 2:</p>
      <fig id="figure-panel-c8d612554fd5a0a24bc121517b6c016a">
        <label>Figure 1</label>
        <caption>
          <title>The rearward situation 1</title>
          <p id="paragraph-121e6af9e9560ea71994e6a5c9e9f348" />
        </caption>
        <graphic id="graphic-e8d6627960add8257ccc3845a8a63198" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_10-41-01.png" />
      </fig>
      <fig id="figure-panel-5341781ff82b43e5079f707fbce74f02">
        <label>Figure 2</label>
        <caption>
          <title>The rearward situation 2</title>
          <p id="paragraph-dea6a9d83991f0d72b86d96ac148c1d2" />
        </caption>
        <graphic id="graphic-445d8adec604ac55e8f22ce67ac15281" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_10-41-28.png" />
      </fig>
      <p id="paragraph-19b5b62bb9753244e401d473ba7fecd1">As we can see, the rearward sense is independent of the source or of the goal of a path, but it has to do with moving in the direction of one’s back (i.e. contrary to the direction one is facing).</p>
      <p id="paragraph-d61fcd61f76696a5b7893b4d5ff422d9">Then there is a sense that we would like to call retrograde. This is a movement or process that happens in the opposite direction to what is usual for the verb or context.<xref id="xref-1f8f33c7fec2c5ff204338dbd3a7276a" ref-type="fn" rid="footnote-ba3a5a595694743ca546c5640ab3dbc3">1</xref></p>
      <p id="paragraph-3">(2)</p>
      <p id="paragraph-4">a. I counted <italic id="italic-018424565101d2d67948888f69977143">backwards </italic>from 5.</p>
      <p id="paragraph-5">b. The movie is told <italic id="italic-c7044e14676b371786575146a19c438b">backwards</italic>.</p>
      <p id="paragraph-87653b0bed65cbc93cc1cee0684f5d16">c. If the scientist was also versed in theology, he or she might see how this cosmogonic event could be construed as a <italic id="italic-22e4764ce4cee69141a3c113c96dd5c0">backward</italic>-in-time emanation from a future “Omega point” that has some of the properties traditionally ascribed to the Judeo-Christian deity.(Jim Holt – Why does the world exist?)</p>
      <p id="paragraph-f474f9297c6aa9c52f95b33b12f606ab"> .</p>
      <p id="paragraph-0cd17d20ffe474d9d3e20202b0c69ef3">We count numbers, watch movies, and move through time in a particular ‘direction’, in the general sense of an ordering of elements in a particular domain, be it space, time, or numbers. If we follow the opposite direction in those domains, then we are dealing with the variety of counterdirectionality that we call retrograde. For human beings, going backward (with the back leading) is opposite to the usual direction (forward, with the face leading), hence every rearward motion is also a retrograde movement. In this way, rearward and retrograde are closely linked logically.</p>
      <p id="paragraph-b491568406e0777c281c4c62acc8d4f8">The most frequent sense of <italic id="italic-6041a8b3d43cecb07f3aca3c2a2b73af">back </italic>is a sense that we call returnative (others call it reditive, Moyse-Faurie 2012). It helps to describe a movement to a former location or counterdirectionally along an earlier path.</p>
      <p id="paragraph-3549c5147a7a76df49b0e06dda7b1847">(3)</p>
      <p id="paragraph-559a0234ed785e94249530889646b362">a. She showered and went <italic id="italic-49c18d9371b3650af08f4d88113f221d">back </italic>to Elysian Park, to Santa Monica and Vine, to 29 Palms. (Salman Rushdie – Shalimar The Clown)</p>
      <p id="paragraph-116c0a4e1af5e01d033a11548ac66d59">b. The soldiers relaxed. Later that afternoon all three Nomans were on a bus <italic id="italic-11f04411d17d4f09328ad38d4243851f">back </italic>home. (Salman Rushdie – Shalimar The Clown)</p>
      <p id="paragraph-69086c10bcc7b439df573caea377afe6">c. I went <italic id="italic-505ac9e4c9d9af45ac9939e09ff0732c">back </italic>a few steps.</p>
      <p id="paragraph-329f01faab3545bf5896fa0bbcf66bb2"> .</p>
      <p id="paragraph-bfa7239fd035642af91ea1bd4424fea0">Suppose a path <italic id="italic-a542451ecccb4d97d47564b1f91296b0">p </italic>is presupposed which the theme has traversed in a salient earlier event and that <italic id="italic-554eae1a2f931ac5f8c60ffe76e82ed0">p</italic><sub id="subscript-1">0</sub> is the source of this path. One prototypical possibility is that the theme follows this presupposed path in the opposite direction, all the way to the source <italic id="italic-34bfac89de9b078c4373acf025d1bb3f">p</italic><sub id="subscript-2">0</sub>. Two returnative senses are possible as weaker versions of this prototype: (i) the theme goes to the source <italic id="italic-465f683c499ccbb3eab2a66b6caf318a">p</italic><sub id="subscript-3">0</sub> via a different path, (ii) the theme follows the path <italic id="italic-ff6275b91d2dfb02f3bbd942034c35f1">p </italic>in opposite direction only partially. We will not further explore these subsenses here. There is a close relation between the returnative and the rearward sense: stereotypically, returning to a former location or along a former path means moving in a direction that one’s back was oriented towards (Allan 1995).This explains how the use of the word <italic id="italic-6638399a45c856d40141f89f628b4866">back </italic>has been extended from the rearward sense (closely related to the nominal <italic id="italic-10">back </italic>sense) to the returnative sense.</p>
      <p id="paragraph-b7d1f22760d6ecd34c99a825591286b4">The following figures aim at clarifying the returnative sense. Figure 3 illustrates what is probably the most prototypical exemplification of the returnative sense, in which one returns to the source of the earlier path taken after turning his/her body and now facing the source. Figure 4 shows another possibility of the returnative sense, closely related to the rearward sense, in which one is moving back to the source but without turning his/her body, i.e. facing the goal but moving towards the source with the back leading. For this sense, what matters is the direction of the path more than the direction in which one is facing.</p>
      <fig id="figure-panel-22b8bbb52bd354c12c9dc77e7ce4f3b5">
        <label>Figure 3</label>
        <caption>
          <p id="paragraph-04eee85fccff144bef785d895d1f4523" />
        </caption>
        <graphic id="graphic-fc6538647ef5681cf253ce49fa8731fe" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_10-47-06.png" />
      </fig>
      <p id="paragraph-996021cd2b6b5776205f8907de07cfce">Closely related to the returnative sense is the responsive sense, indicating that an action from A to B is the response to an earlier action of B to A.</p>
      <p id="paragraph-25cd04384d8fbf8ab6cdc3ff241d6806">(4)</p>
      <p id="paragraph-44c699ac075a44c7cc083187f759e3f9">a. Her mother was calling to her from the far side of the globe. Her mother who didn’t die. Kashmira, her mother called, come home. I’m coming, she called <italic id="italic-f0e3610fa9f1dc2a8a941fbbc995a42b">back</italic>. I’ll be there as fast as I can. (Salman Rushdie – Shalimar The Clown)</p>
      <p id="paragraph-a87b7abf93f1489b17efa94136ac5565">b. A boy kicked Quentin in the shins, but he didn’t kick <italic id="italic-b8cebc9401ccf5520ad5093568e8d941">back</italic>.</p>
      <p id="paragraph-8f90e69970b5aa2efd2174fbb009f952">.</p>
      <p id="paragraph-fdc5435bd09d0a5a1a6db799aef40df9">It is Kashmira’s call in (4a) that is ‘going back’ to her mother (not Kashmira herself) and it is a kick of Quentin in (4b) that is denied to have ‘gone back’ (not Quentin himself). We can understand the responsive sense as involving abstract paths from agent to patient along which the action goes. The counterdirectionality of the responsive sense is no longer defined spatially, but a previous path of action from B to A is followed by a path of action from A to B. In other words, there is an abstract notion of directionality that is shared between motion (from source to goal) and action (from agent to patient) and hence between the concepts of returning and responding, which closely relates the returnative and responsive senses.<xref id="xref-5d9a97e9ec007aba4b07f962a9ddcce0" ref-type="fn" rid="footnote-b877bcaa5aabc7c10c8dd7227b7700cf">2</xref></p>
      <p id="paragraph-27d883510af1fd5f6b955ce57e783c6b">It is a relatively small step from the spatial domain to the domain of change. This is where we find the so-called restitutive senses of <italic id="italic-71a22b72eae7e43c858fe68646f9c407">again </italic>and <italic id="italic-28b8dcb02cf1e357a4916a301ddcad68">back</italic>.</p>
      <p id="paragraph-1bddd277985abf56080e7be203ad7cb2">(5)</p>
      <p id="paragraph-4cbf8c48c9b77356c3f8bfc2316a5742">a. “Alas, there are many who do not love God in these days,” Lady Yvolde said, “but we must pray to God that He will set the world right and bring men <italic id="italic-81e39cd1004d13777ce39dcef91b23af">again </italic>to virtue.” (Connie Willis – The Doomsday Book)</p>
      <p id="paragraph-60ffae0151ae7d1ba6420468aaca512e">b. The noise jarred him <italic id="italic-3ce45599743e2552259487219a6dd0c9">back </italic>awake. (McIntyre 2012)</p>
      <p id="paragraph-57cec7bec7f8fc003e3a7c8b6d9fe362">The change can be along a continuous scale or in a domain of discrete values (e.g. ‘asleep’ and ‘awake’). In both cases the change can be seen as unfolding directionally along a path, in a non-spatial domain. As a result, motion and change share general properties of directionality and therefore also the returnative and restitutive concepts.<xref id="xref-4f685a4fcc584303c003060df8e2996b" ref-type="fn" rid="footnote-5b3ba0fe95173e1f9c9c6c9f3a0d92d2">3</xref> <italic id="italic-220395ebe346ab1aad6db06a3e8c0f76">Again </italic>is typically used in English here, but <italic id="italic-496bad2f914ceda6ff15c3466f1e3e09">back </italic>is also possible (Beck et al. 2009, McIntyre 2012).</p>
      <p id="paragraph-17ff29649a76319ed149eaef0c5ccd42">Finally, we arrive at the repetitive sense, expressed with <italic id="italic-94fbace38377f04c045b9f37116f61a3">again </italic>in English, or with <italic id="italic-e300ceb9bd7cb446d708172095c616ff">once more</italic>, <italic id="italic-a0ed376ad8364a1564ab9d8d37c18528">once again</italic>, or <italic id="italic-92fd9ef1bb182254bd258acd91963de7">anew</italic>:</p>
      <p id="paragraph-9c910031cc28e28d9fc7d988ed84a297">(6)</p>
      <p id="paragraph-55ea1b21092b87934abee9918288dfc6">a. He stopped in midsentence. “Without what?” she teased him, guessing the unspoken words, but he blushed <italic id="italic-7aea2370470a83e61ca0db7973d90a1e">again </italic>and concentrated on the road ahead. <italic id="italic-9e20e74162501828c73669612a58d71a">Without a woman’s touch. </italic>(Salman Rushdie – Shalimar The Clown)</p>
      <p id="paragraph-d3fab1581da51558b039999be06274ba">b. “The trial is <italic id="italic-49f8132fd34519689a3ce778e255016e">once more </italic>delayed,” Kivrin heard Gawyn say. “The judge who was to hear it is taken ill.” (Connie Willis – The Doomsday Book)</p>
      <p id="paragraph-be63dae865460164457b3bccad00f855">c. The nurse led him out. “If he wakes and calls for you <italic id="italic-3720496aefd09dba8dc9b8ecb3b0b09d">again</italic>, where can you be reached?” she asked. (Connie Willis – The Doomsday Book)</p>
      <p id="paragraph-10">d. The door opened <italic id="italic-11">again</italic>.</p>
      <p id="paragraph-e7362a31c55f9cdb5a89d90ceb6ac67c">.</p>
      <p id="paragraph-12">In this case the existence of a previous event of the same type as the described event is presupposed. Since <italic id="italic-12">again </italic>can be used both for the restitutive and the repetitive sense, a well-known ambiguity can be seen in (6d): in the repetitive sense the door had opened before and that change of state is repeated, and in the restitutive sense the door was open before and that state is restituted. There is obviously a close relationship between these two senses, because the restitution of a state implies the repetition of a state. This has fact led people to an analysis with one repetitive <italic id="italic-13">again</italic>, that depending on the position in a syntactic structure, can lead to repetitive or restitutive interpretations (e.g. Von Stechow 1996). For our purposes it is sufficient to recognize that in a restitution there is always a repetition. If the door had been open before and later on there is a change from shut to open (a restitution), then also the state of being open is being repeated. This motivates a link between the restitutive and repetitive senses. See Pedersen (2014) for more discussion about the inferential relation between repetition and restitution.</p>
      <p id="paragraph-fe30b794ef41a7961363e69743caf0c0">We can summarize the senses we have discussed in this section as a network or map in the sense of Haspelmath (2003) and Zwarts (2010):</p>
      <fig id="figure-panel-522ff65aac7c900ac565548f59660587">
        <label>Figure 4</label>
        <caption>
          <title>The network of counterdirectional senses</title>
          <p id="paragraph-6ae1fdcf76cee8e7aec419fd203251bf" />
        </caption>
        <graphic id="graphic-ccc6d79c2dbfcaa1c8db2dcbc521c6f2" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_10-55-29.png" />
      </fig>
      <p id="paragraph-0d91e85fa5200cf5abe93e277213bbdf">In this network the lines represent different ways in which the senses are related, but always in such a way that the senses that are most closely related are connected by a line. Because of this, we see something of a coherent domain appearing. Different languages express these senses in different ways. We have seen how English uses <italic id="italic-4e300ea9e7119a3b940ffb51d7e2b288">backward(s) </italic>for the rearward/retrograde corner, <italic id="italic-bd54b4654d0bd1df99943a193c3bdf9b">back </italic>for the returnive/responsive/ restitutive part, and <italic id="italic-7b882207e8477f74bafd334b704853d9">again </italic>for the restitutive/repetitive side.</p>
      <fig id="figure-panel-2f912d5beac45e9674ca8d0ef270ef58">
        <label>Figure 5</label>
        <caption>
          <title>Lexicalization of counterdirectional senses in English</title>
          <p id="paragraph-d81a7af8ccb991cc38a41009fa1287ce" />
        </caption>
        <graphic id="graphic-a4ecef4012541e5514ebc4abb173140b" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_10-57-51.png" />
      </fig>
      <p id="paragraph-9a30ebad892669cb0144285c227f9859">The map allows us to hypothesize that the set of senses covered by <italic id="italic-fcbcf1221eec8a2ecabdc4eef11cd2f9">back </italic>has developed from the etymologically primary rearward sense, closely related to the notion of the human back (Allan 1995), to a range of more abstract senses.</p>
      <p id="paragraph-ce600ec3a09d41189743f0fa7ee4c05d">Much more discussion about and refinement of these senses is necessary and possible, but we think that this map offers a good basis to turn now to a detailed exploration of the data in Brazilian Portuguese and Dutch.</p>
    </sec>
    <sec id="heading-1">
      <title>2. Counterdirectionality in Brazilian Portuguese and Dutch</title>
      <sec id="heading-2">
        <title>2.1 The rearward sense</title>
        <p id="paragraph-c202081d1451d4322cbfbc55d4697d6c">We are dealing with the rearward sense when a person, animal or object is moving in the direction of its back.</p>
        <p id="paragraph-a1ea6c7a73de469b95d28e2977a0df9f"> In BP the rearward sense is usually expressed with the verb <italic id="italic-78b197e5e3c5fc610973df207b4cae37">recuar </italic>or the adverbials <italic id="italic-58a92e28e92298fda2332ed5caf4a9de">para trás</italic>4<xref id="xref-5cdae468953ff854a097ad86e2351e68" ref-type="fn" rid="footnote-799c528c9f5c041c0439dd6e76e4c0a3">4</xref> and <italic id="italic-51509469a01f78eff7739104dcce9eb1">de costas </italic>when combined with verbs of movement. It is important to notice that although <italic id="italic-b830218a2a9097e55c6d4ed0bb870a1b">recuar</italic>5<xref id="xref-3ea55d428da0fb4df8e75e0b9934784b" ref-type="fn" rid="footnote-e8bce4dac02d3d44e931918744ada1e9">5</xref> begins with <italic id="italic-f48f046a1c74282fd943c012ac2ff317">re-</italic>, one can argue that we don’t have a ‘re-’ prefix in this case because there is no verb <italic id="italic-fef0bffe9a9bf933e9ac6672018f1f7f">cuar </italic>in BP.<xref id="xref-f377380b77884f67665122edcdb0a18f" ref-type="fn" rid="footnote-b0c1e335e24c911deff3abb9eb6d0230">6</xref> <italic id="italic-10a7fde9d7ced9ab9777addb8fa50183">Recuar </italic>can be translated into English as ‘retreat’ or ‘recede’, and it always implies rearward movement, but not necessarily a movement towards some source.</p>
        <p id="paragraph-c7f1e44230393f999bd60ab743afb610"><italic id="italic-52bb5e1bdcd0aa2d86659f7becf282e6">De costas </italic>indicates backwards (or contrary to the usual) direction, for instance, when somebody swims <italic id="italic-7ca1a56f6a328bf7e5453708c9f680d4">de costas </italic>her back is in the water but not her belly, and when somebody sleeps <italic id="italic-c560458e618f97c49f25764b0d61c26c">de costas </italic>her belly is on the bed but not her back. Finally, when somebody walks <italic id="italic-58659239fa63146ff273ad27d30e020a">de costas </italic>it means the she is moving in the direction of her back but facing front, as in the following example:</p>
        <p id="paragraph-d17db2d2c70f76be92b683e0dc0da5b9">(7)</p>
        <p id="paragraph-00ddf5f2ae8eb0987b4ff8c68a9786ce">Indiano anda de costas por 25 anos e esquece como é andar para frente.<xref id="xref-b4c666e8882d102ed7b5f559702195db" ref-type="fn" rid="footnote-10d15baec9113ac81544071f6e239f87">7</xref></p>
        <p id="paragraph-0ba44312ed044e0b418528c3cb4724ff">Indian person walks backwards for 25 years and forgets how to walk forwards.</p>
        <p id="paragraph-726ad6ad5b646af2d692624c59226ca9"> .</p>
        <p id="paragraph-00c365a82810cafbcd01b9c890c64031">The adverbial <italic id="italic-31a823847cebd2d384c13c0dae70fbe6">para trás </italic>can also be used to indicate the rearward sense, as in the following examples:</p>
        <p id="paragraph-9335bdf689ffdcfc3662cf083fb02399">(8)</p>
        <p id="paragraph-d65f5f18a1ba9e025a602672d5e0c8c8">a. Cair para trás</p>
        <p id="paragraph-86d8f581f353c9284395e3c4a307c77b">‘to fall in the direction of one’s back (backward fall)’</p>
        <p id="paragraph-2d372f395ba374cbe9463c44a41cef65">b. Andar para trás</p>
        <p id="paragraph-6a0db2eae6200a60b777de7f9177a766">‘to walk in the direction of one’s back without turning’</p>
        <p id="paragraph-13"> .</p>
        <p id="paragraph-14">The expressions <italic id="italic-c06cad4cf8bbd3d6b171291c5eb45cf7">de costas </italic>and <italic id="italic-44dabcd8b82666e524c3f476bc52c952">para trás </italic>have a wide range of uses and here we’re considering only their so to say spatial-movement interpretations. It’s important to note that <italic id="italic-d42214166eb9439064f54149cd5e4045">para trás </italic>has also a retrograde sense, as we’ll see in section 3.2.</p>
        <p id="paragraph-15">(9)</p>
        <p id="paragraph-16">a. João recuou.</p>
        <p id="paragraph-17">João went back (leant back)</p>
        <p id="paragraph-18">b. João andou de costas, mantendo contato visual com seu irmão.</p>
        <p id="paragraph-20">João walked back (in a backward manner) keeping eye contact with his brother.</p>
        <p id="paragraph-6d83bd4635a413ce310a62d9f704fad7">.</p>
        <p id="paragraph-bb76a109f61c03dede5a47f608466838">In Dutch we find two adverbs: <italic id="italic-d627e156dcae005e6e7005b9312f6017">terug </italic>lit. to-back ‘back’ and <italic id="italic-47e3636e1df6e4e1b697a13ca03945a1">achteruit </italic>lit. behind-out ‘backwards’.</p>
        <p id="paragraph-10698a5d6ac0365f55468e5e1e34ce18">(10)</p>
        <p id="paragraph-4b3f23cb6c9846a12b0297638dd1272f">a. Jan deinsde terug.</p>
        <p id="paragraph-b3c17a0026df9294b8bea3a1f6e1a164">Jan shrank terug </p>
        <p id="paragraph-a8705efdc93d96bc825163623898e40d">‘John shrank back’</p>
        <p id="paragraph-6f6e05959c073ce98274bbdf5c739ee9">b. Jan reed achteruit.</p>
        <p id="paragraph-a04207f2bc87575aabff49f6888ef401">Jan drove achteruit </p>
        <p id="paragraph-ef49391996a4a1fe68029a71c44adb85">‘John drove backwards’</p>
        <p id="paragraph-9a54b6fbb8aef72fcb7f905de3b2516c" />
        <p id="paragraph-9e4844c32230a463215650714768b834"><italic id="italic-814a0d2d1d009bd886248a490cebe1f4">Achteruit </italic>is the productive adverb, possible with an open ended range of verbs, while <italic id="italic-24f65d465f3508974f88779fe6b20fa5">terug </italic>is restricted to a small number of verbs, often with special meanings (e.g. <italic id="italic-5cbb4cabaa058bae9fb8325982a638ca">terugslaan </italic>lit. back hit ‘backfire’ of a gun). The verb root <italic id="italic-289370187ed7b673e93e0008ac08bd69">deinzen </italic>is a ‘cranberry morpheme’, it does not occur apart from one of the rearward adverbs <italic id="italic-8abf6acddd5288cce5a75980d1c90809">terug </italic>and <italic id="italic-47c1d32da1c1d91b5315a6325cb24ce0">achteruit</italic>.</p>
        <p id="paragraph-cfc1f6fba20b8fe127494857b9a386e5">As a summary, what we have here for the expression of the rearward sense is two adverbs for Dutch (<italic id="italic-3d536f1a90712e5a68857c6c4105286a">achteruit </italic>and <italic id="italic-66499d41bbd85fec93f591db36c28234">terug</italic>), and for BP some verbal forms (such as <italic id="italic-542b90f508bbdf77997301841a126d03">recuar</italic>), and adverbials such as <italic id="italic-7ea74e5da61fcdde0cfc5280e7b953a5">de costas </italic>and <italic id="italic-119d682a9e0943ff3516442040fa19e4">para trás <italic id="italic-14"/></italic>– the first one of them more specialized for the rearward sense.</p>
      </sec>
      <sec id="heading-1d2a8594c986dd4ea08f0ac8974c8e5b">
        <title>2.2 The retrograde sense</title>
        <p id="paragraph-a49d9a22d816f1742131769bb6aba7e4">Another area where we see counter directionality is when a movement or process happens opposite to the usual direction. This could be seen as a generalization of the rearward sense, as we said earlier.</p>
        <p id="paragraph-d322937616a46f09ccedde01438a2e4e">First BP. In this language, we find the expressions <italic id="italic-15">de trás para frente</italic>, <italic id="italic-16">para trás</italic>, <italic id="italic-17">ao contrário </italic>and also <italic id="italic-18">re- </italic>forms.</p>
        <p id="paragraph-03d1c5d2887a5a0f4b431f8af7e3efd3">(11)</p>
        <p id="paragraph-96728cb13fdbe3784c4a14b07d9dd883">a. Você consegue contar de 0 a 100 de trás para frente/ao contrário?</p>
        <p id="paragraph-c020dda0e520a251aebf1645ffb5b3ff">Can you count from 0 to 100 backwards (i.e., from back to front)/on the contrary?</p>
        <p id="paragraph-2695377cd84b10a108199ce1a3545c88">b. pedalar para trás / pedalar ao contrário </p>
        <p id="paragraph-7c0ef3573ea13d7ff6b0f9263eec9f35">to backpedal (i.e., to pedal to the back) / to pedalon the contrary</p>
        <p id="paragraph-f2905a34f828d518136904efd9ab4aa6">c. virar para trás / virar ao contrário/ reverter<xref id="xref-61b7c4ab8786268a1f6de2ff5c03571d" ref-type="fn" rid="footnote-2a4d93373603971d3a2728289d748f63">8</xref></p>
        <p id="paragraph-238ff4fef737dd3280e7322b2d8266ea">to turn back / to turn on the contrary / to reverse</p>
        <p id="paragraph-bc22e0a24c6f04f3dec44fa493caac66">The expression <italic id="italic-c9fcf0f596ea630ed180c536ab644187">de trás para frente</italic>, which literally means ‘from back/ behind to front/face’, implies some sort of pre-ordering and it indicates that this ordering is being followed backwards, such that the end is now the beginning. This is particularly clear in the example (11a) in which the natural endpoint (100) is now the beginning, and also when someone follows a recipe <italic id="italic-1acef8d4ab8ec0efe8a1e9fcc7bb5e7f">de trás para frente</italic>, beginning with what is supposed to be the final step.</p>
        <p id="paragraph-f26025776ba9a531084741492fce9850">BP also has the expression <italic id="italic-b372d651ab7a151078b7dc800dd27905">ao contrário </italic>(‘on the contrary’, ‘contrariwise’) which has a wider range of uses and it simply indicates that something is being performed in the opposite of the natural way; that’s why one can backpedal (<italic id="italic-dc68c315c568f16eab2786fd74f562e9">pedalar ao contrário</italic>, i.e. pedal contrariwise). It’s possible also to find <italic id="italic-1a8b67e93a2b2254fe7eaacfb871f000">para trás </italic>for the expression of the retrograde sense, which confirms that the rearward and the retrograde senses are connected.</p>
        <p id="paragraph-aa55074d7f6ca80f312e28284439d2fb">In Dutch we find here <italic id="italic-23f15b0df95e65035c59688f7bb41b6e">achteruit </italic>and <italic id="italic-0a4cf30ac1323d5b2645fb70297a632e">terug </italic>again, but also a more complex expression like <italic id="italic-74a2be4813a87311ddee4f2815203e9e">van achternaarvoren </italic>‘from back to front’.</p>
        <p id="paragraph-3d80469df7925376f8a2dce53347369a">(12)</p>
        <p id="paragraph-d2ea1415be68f07284d6def0f7ed65a0">a. Kun jij van honderd tot één terug tellen?</p>
        <p id="paragraph-2032123e4d21e1447dffc0f7937954dd">can you from hundred to one terug count</p>
        <p id="paragraph-ad2dc2f0a376be46273b8e9ee27459d0">‘Can you count backwards from 100?’</p>
        <p id="paragraph-9a79988f500eb67c12c65a14eeaacda6">b. Een palindroom kan van achter naar voren gelezen worden</p>
        <p id="paragraph-c6271365bfd070f66951ab97eed20197">A palindrome can van achternaarvoren read be </p>
        <p id="paragraph-2fc1fc04331b2a0c573a8004b45f6317">‘A palindrome can be read backwards.’</p>
        <p id="paragraph-427cf2f803f6077d4d5dd4ef31230515">c. De klok gaat morgen achteruit</p>
        <p id="paragraph-1140d53739ab8bd7ca6f4bc0b4dd4f13">The clock goes tomorrow achteruit</p>
        <p id="paragraph-a689e37c99d24b26b9b455794f19d43b">‘The clock goes back tomorrow.’</p>
        <p id="paragraph-fd33a5e7025006d5641de34d1b024782">We can conclude that a lot is still unclear about the precise conditions under which one expression or the other is chosen for the meaning that we call retrograde. Collocation and lexicalization might play a role here. <italic id="italic-a795e0ff0617e82b6c847647059e35ba">Terug </italic>only seems to be possible with certain verbs (like <italic id="italic-3146a91397646369ca900d822c165ec3">tellen </italic>‘count’), while <italic id="italic-e1de62672b7980515c7f4000f3b5ffc5">van achternaarvoren </italic>and <italic id="italic-9298c97327062092f530a4a04035f627">achteruit </italic>are more productive. The choice between <italic id="italic-6d86ba51d81209d4addc332f33ce19e2">van achternaarvoren </italic>and <italic id="italic-9e1544f19c96093c8651014039c40564">achteruit </italic>might be determined by additional features, such as the dimensionality and directionality of the underlying domain. A word is a one-dimensional object, in a sense, with a front and back; a clock has a more complex dimensionality, and although there is a clear directionality in the way the hands move (‘clockwise’), there are no front and back sides to it.</p>
        <p id="paragraph-4fb7a6800421128cc5e9c4c0798152c6">The specialized expression for the retrograde sense in BP is <italic id="italic-955a0dd010c9f24434e82375af48182e">de trás para frente</italic>, alongside with expressions such as <italic id="italic-790c075f99bb807fc5d40f095d7f1e25">ao contrário </italic>and <italic id="italic-9c0830f4ded05a4e87b70787da2bcb39">para trás </italic>which have a wider range of uses. For Dutch what we find are again the items <italic id="italic-97ca2e8da4ef836bc9ad391984c9e1ac">achteruit </italic>and <italic id="italic-44064d656bb47b4f6596cba631e8cabe">terug</italic>, and the more complex and specialized expression <italic id="italic-75029eb54081e4ffa9bcfa387e5aa80d">van achternaarvoren</italic>.</p>
      </sec>
      <sec id="heading-4d5289e566c88c7fea252e9aad0da176">
        <title>2.3 The returnative sense</title>
        <p id="paragraph-469d4a32fcc76785a7eae01a9acb0538">The returnative sense involves a presupposition that the theme has been at an earlier position or has made an earlier movement; the theme is going to the position where it has been before or it follows a path that it has gone before in the opposite direction.</p>
        <p id="paragraph-0fa0554e06460a7e7a2d818d1731af3d">For the returnative sense, in BP we find the adverbial expression <italic id="italic-7194e21f933c310d07570e9b6c807f1e">de volta</italic>, and the verbs <italic id="italic-193f2ade6b5ab11ffc2587b4fa922f1f">voltar </italic>and <italic id="italic-a6fe5e7bc8f7a6b6375826436afdc426">tornar</italic>, and also constructions with the prefix <italic id="italic-07426a9609c229b7640b842f90088f3b">re-</italic>, which are usually more lexically specialized.</p>
        <p id="paragraph-abb6cb9824747c7b3c67fa6139d204e2">(13)</p>
        <p id="paragraph-b48bb4a1f890cdc7cc96f63aae2b495e">a. Os convidados voltaram / retornaram para casa.</p>
        <p id="paragraph-65c4c0dc980a64914f1a341ddf1195fa">The guests went back/returned home.</p>
        <p id="paragraph-aa80fa474e6327aca2b23bf4dd8776fc">b. pôr de volta / recolocar</p>
        <p id="paragraph-a62bc401c7eee691b8d65bd685a72702">to put back / re-place (put something back in its original place)</p>
        <p id="paragraph-b5e3716ef019bbdd45d9b6e02d041962">c. devolver / retornar / dar de volta</p>
        <p id="paragraph-19ef3c13732620ffb03e472752a23b43">return (devolve) / return / give something back</p>
        <p id="paragraph-f89ab8974caeb59bb296488a8349b5d0">.</p>
        <p id="paragraph-db5513eb1250258e418911225bbd504d">In BP, the verb <italic id="italic-60296199124aebf8f96202a3897a7c7d">voltar </italic>and movement verbs prefixed with <italic id="italic-addc177042665c11497975f777a9987d">re-</italic>, such as <italic id="italic-e6599fda1f2746de87b59f15cb0c584b">retornar</italic>, are the common choice to express the returnative sense, which can also be expressed via constructions with the noun <italic id="italic-5e32b852d4063f371d4e62618d56f041">volta</italic>, in expressions such as <italic id="italic-88e4ef52d2b235f652f07b7ab259ed83">de volta</italic>.</p>
        <p id="paragraph-edfe76501668cecb9f9746dca2e80fc5">The form <italic id="italic-c462e91a4af4368895d5266a9ec9ec10">voltar </italic>is directly related to the Latin form *<italic id="italic-6b438d378f60ce6538592e3c8f8d0dac">voltare</italic>9<xref id="xref-59d083c3d8ea28ef46fe7c20f7fe5aa3" ref-type="fn" rid="footnote-5de877820a4606f36e8c9241b541d6b6">9</xref>, which basically means ‘to roll’, ‘to spin’, ‘to revolve’, and nowadays in BP its basic meaning involves something or someone going back to its source or initial position. There are also some options with the prefix <italic id="italic-09dea2a30a25c55ccaa38f5bf526ce24">re-</italic>: <italic id="italic-d00f7a8b02f77333f7fa8fb5159e4459">re- colocar </italic>(which is transparent for re-put ‘put back’or ‘put again (in the original place)’), and <italic id="italic-c22b618851ed2b634346d023deb004c8">re-tornar </italic>(which is not transparent, since the verb <italic id="italic-72ea37642d342950cd71057262c87714">tornar </italic>means ‘to turn’). As we saw with the other senses, usually the <italic id="italic-5620da9316d287e3c612736ce98e45c9">re- </italic>forms are more formal, probably because they are lexically specialized.</p>
        <p id="paragraph-54188624ed757f79781685beb7a16c90">In Dutch we only find <italic id="italic-cf07d2c0f64b61dc7f10679bcf6c09a7">terug </italic>here (although older Dutch still had <italic id="italic-0b266ed9848ac909afa181833dbd26ef">weder <italic id="italic-5c836c98b667b5d1b2482394e7d78708"/></italic>here or its contraction <italic id="italic-be670f8ef1f516a4a6163f6bc18b0078">weer</italic>, the cognate of German <italic id="italic-4713c888580b7fbc8535285d7b228e79">wieder</italic>).</p>
        <p id="paragraph-eaf1021f902b5467f2d999e8beeb3603">(14)</p>
        <p id="paragraph-df02513bbccaf6b34ab4d1618267a189">a. De feestgangers liepen naar huis terug.</p>
        <p id="paragraph-7f98c53ac62e0f4f908735760c267562">the party-goers walked to home terug</p>
        <p id="paragraph-0ccea0710b67a4de0e508cc04bc14c17">‘The party-goers walked back home.’</p>
        <p id="paragraph-2e0edca4a828216d810913048a540297">b. Jan is terug.</p>
        <p id="paragraph-f8283e98017eab31f65f64083c54f702">Jan is terug</p>
        <p id="paragraph-038bec32e76660b1fee0fd702d79ce1b">‘John is back.’</p>
        <p id="paragraph-9a31445ce10c17ab74040bba7174785a">c. Je moet het boek terug leggen</p>
        <p id="paragraph-4ca1911a8dca316669923c408a7437a1">you must the book teruglay</p>
        <p id="paragraph-46f1f19953cb825778fb21295983b981">‘You have to put the book back.’</p>
        <p id="paragraph-4e61e6fee66c45b34871eec12d88430a">d. Ik wil de auto terug geven</p>
        <p id="paragraph-91ec9f8caa28cab4a8e2a294e4cc5121">I want the car terug give</p>
        <p id="paragraph-fd51191f40f66b63db2e6f4e0646a3d6">‘I want to give back the car.’</p>
        <p id="paragraph-43851c3c1af3f5765fd96a0d5e1b5fe6">.</p>
        <p id="paragraph-809cc2c94bb36cd24f125a6010dcab5d">In both languages, the options for the returnative sense seem to be restricted – <italic id="italic-667a1cd743166baaf28afe0263532367">terug </italic>for Dutch and constructions based on the verb <italic id="italic-6448d00c5f0c37828b7252636f3cc170">voltar </italic>in BP. It is interesting to note that Dutch and BP have different etymological sources for the returnative sense. While the Dutch returnative expression derives from the word for the posterior body part (<italic id="italic-af632ad4c1ee652bd1228b452eaa8c23">rug </italic>‘back)’, the BP returnative expressions derive from roots describing a rotation (of the human body). This might be a more general difference between Germanic and Romance languages, witness the English word <italic id="italic-d5c7fc0ce50e9267abc5f1a4d8eb3d85">back </italic>and French <italic id="italic-c43db2bfc67ca99da99c6dcce84bb091">retour</italic>. As we’ll see in the next section, the returnative sense is closely connected to the responsive sense.</p>
      </sec>
      <sec id="heading-bd6aa5571052ea941b472db42aae2ca3">
        <title>2.4 The responsive sense</title>
        <p id="paragraph-21396c321547f4468e1dfe5f039d325f">This sense can be seen as a thematic extension of the returnative sense, as we saw in section 2.</p>
        <p id="paragraph-ddbd84d78f34b52a68d7f154494f4d0f">Again, for BP what we find is <italic id="italic-7f0adc1e54efbd9fc2d30b246eebef7e">de volta </italic>and some <italic id="italic-aca23e03581b13caf4f31a8a7f6502b7">re- </italic>compounds:</p>
        <p id="paragraph-f6453a099b1f32a25fd215b03c6a9e48">(15)</p>
        <p id="paragraph-574714491bf64a0891a84e1623d2eb74">a. Ela escreveu de volta, dizendo que estava vindo.</p>
        <p id="paragraph-032dab8a88e13d9b94dd9e33b1030ec1">She wrote back, saying that she was coming.</p>
        <p id="paragraph-9426abec8b7c2b8bbeb46a5e07a96b00">b. Ela respondeu / retornou [<italic id="italic-95d217c27af51a71895c1b974df42975">more formal</italic>], dizendo que estava vindo.</p>
        <p id="paragraph-535707d98b4452d413bf6462fce20ac7">She answered / returned, saying that she was coming.</p>
        <p id="paragraph-5fc70c6f42c801079aab438c5149290b">c. chutar de volta / devolver a bola </p>
        <p id="paragraph-0b60a39bc0007180793c015f0b8493d1">to kick back / to return the ball</p>
        <p id="paragraph-c206a7aba7e17d7b6f7eeabab4a0a6bb">d. responder</p>
        <p id="paragraph-be68d705f83ed98337594e515cb6fcd5">to answer (back)</p>
        <p id="paragraph-0b55b8b62c81d7871a44f7fbff24a25d">.</p>
        <p id="paragraph-93da92c40499fbc0544125b4bf180513">In Dutch we find <italic id="italic-7bda3c1ab2511b274e70c3e4c1943176">terug</italic>:</p>
        <p id="paragraph-8fa50d7e5b751532d7015e06b272b621">(16)</p>
        <p id="paragraph-a21e32db7035c39a6682b8e42ec1513e">a. Zij schreef terug dat ze kwam.</p>
        <p id="paragraph-f699e4bdce1a618933ff4999fcbfbbda">she wrote terug that she came</p>
        <p id="paragraph-074e903af70cf98e83842f37b94836f9">‘She wrote back to say that she was coming.’</p>
        <p id="paragraph-b9bfa6d776d08b1d51add961fbda7299">b. Hij schopte de bal terug. </p>
        <p id="paragraph-cc7f409f6a3daed45069c1631dcda579">he kicked the ball terug</p>
        <p id="paragraph-08b4bbbbe85b5f31cb80cd668bab4ab5">‘He kicked the ball back.’</p>
        <p id="paragraph-d6c50aeb1bc41739991ff5708aa81fe3">c. We moeten de buren terug vragen na onze visite.</p>
        <p id="paragraph-ddef44aa7c9e2b5b84918244ccd6ff23">We must the neighbours terug ask after our visit</p>
        <p id="paragraph-7370a08bb042128552d39e29d590d778">‘We must return the invitation of the neighbours after our visit.’</p>
        <p id="paragraph-42ed0203767367471bf658c40d613642">.</p>
        <p id="paragraph-efdb05102fbaa43757ef1765c2f29cc3">In each case <italic id="italic-d4c4faf37a2fbf5a89fe8685654c107c">terug </italic>triggers a presupposition that there was an earlier event in which the subject received a letter (16a), the ball (16b), or an invitation (16c) and that the sentence describes a response directed towards the agent/source of that earlier event.</p>
        <p id="paragraph-4986992089125cebb963c3e35932c3c0">Basically, what we see in both languages is that the verbs or adverbs that are used for the returnative sense can also receive a responsive interpretation. This might suggest that the responsive sense is a special case of the more general returnative sense, rather than a special historical development.</p>
      </sec>
      <sec id="heading-74cf7d9b8f85b2b78e876f1f1216bfa6">
        <title>2.5 The restitutive sense</title>
        <p id="paragraph-5ebac566b5043bba07296f86148538a7">In the restitutive sense an earlier state is restored or an earlier process is reversed. Consider the following sentence in BP which expresses the restitutive sense:</p>
        <p id="paragraph-c2926a908f660f0b3fe7e7e20137ed05">(17) </p>
        <p id="paragraph-ce8cba27909b6f4fd9628cbd6d843316">João dormiu e acordou de novo /novamente/ mais uma vez / outra vez às 2 a.m.</p>
        <p id="paragraph-12517431b059c9f502fb757ee2b4ccdd">João slept and woke up again / one more time / another time at 2 a.m.</p>
        <p id="paragraph-4a904a3720a17c2e4ec5b171301a59f1">.</p>
        <p id="paragraph-c8f6eb56da7e5a84f164ab6d6a82e21f">In BP we find the expressions <italic id="italic-f37005274e9c6818442d783d07d0de15">de novo</italic>, <italic id="italic-64a5b8e2706cb981389a52dbf5e92d24">novamente</italic>, <italic id="italic-d38cc965b207035daacbab7288f57400">outra vez</italic>, and <italic id="italic-4529ddeea0c1992b25c215b3509282d8">mais uma vez </italic>here, where similar expressions in English, German, Dutch are only used for the repetitive sense.</p>
        <p id="paragraph-0d001ab432794fb6e1daaa28da5a6003">In standard Dutch mostly <italic id="italic-a3145b3339ce1382a264c65eeb5ed72d">weer </italic>is used here, but <italic id="italic-bb9e4f0ed94e16b29bd40baedfb2971e">terug </italic>is possible with certain verbs and in non-standard Belgian Dutch <italic id="italic-eb4d9e0686e4b558ad5e94b5a63040e4">terug </italic>is used much more productively for the restitutive sense. In addition we find the prefix <italic id="italic-f8c7951ead7350b233385b16b20dd842">her- </italic>‘re-’.</p>
        <p id="paragraph-9bf7aab02f272924fe4975a7ff7c0bae">(18)</p>
        <p id="paragraph-ad273ba0e6b1eb022f56d45a51e15449">a. Hij werd terug/weer wakker.</p>
        <p id="paragraph-54d1e25c2a7ae9a156dbb602ac5acf79">he became terug/weerawake</p>
        <p id="paragraph-ff7f4b13b6dca171a0cd1db4780ff3de">‘He woke up again (after having fallen asleep).’</p>
        <p id="paragraph-c1596225a785c1239e308b1e78e86552">b. De stad moet terug/weer worden opgebouwd.</p>
        <p id="paragraph-ec215ee34c2ccbefe4747cb444709481">the town must terug/weerbe build-up</p>
        <p id="paragraph-ed6259c98133ed869fedbece3b73c128">‘The town must be build up again.’</p>
        <p id="paragraph-a2f35887a12a0afe9465353ae55f5c22">c. Het deksel kan eenvoudig geopend en terug/weer gesloten worden.</p>
        <p id="paragraph-ab2ebbc158cfa9926e361601b8a91975">the lid can simply opened and terug/weer closed be</p>
        <p id="paragraph-00149a7ad8e6cf4d4f24a4815b745b21">‘The lid can be opened and closed again in a simply way.’</p>
        <p id="paragraph-cfc0709b22f0a9f0243cf20e4eddf917">d. De stad werd heroverd.</p>
        <p id="paragraph-8ffb36ef62eafb075ac6240afbcb9eb4">The town was her-captured</p>
        <p id="paragraph-51e71071bfe446fcfa204700e4a847dd">‘The town must be recaptured.’</p>
        <p id="paragraph-f8d79946072ed1a0dee82f5bd0ad5ddd">.</p>
        <p id="paragraph-53ca1ddc679181ec299b2240a63aff5e">Concluding, BP has a number of expressions to mark restitutive change: <italic id="italic-b428975e7bdb8527d21f573cd18fe7f9">de novo</italic>, <italic id="italic-df11bfeb5a5b27b23f7d0607297e08f7">novamente</italic>, <italic id="italic-8ddce0038aa061262d77aaf527c7a906">outra vez</italic>, <italic id="italic-7f04d197bb22b79e813c5d57c3e80aa2">mais uma vez</italic>. Dutch has three forms: <italic id="italic-ccc96c6f7f6658ace3f6b00f913b5b6e">weer </italic>and <italic id="italic-7163b3aed2540715f0aea211beebb24b">her- </italic>in standard Dutch, and <italic id="italic-60dfc35da3f10466dad356a1dc7c8674">terug </italic>in non-Belgian Dutch. None of the three Dutch forms are specific for the restitutive sense, but they are (<italic id="italic-febed8eeecf464dff0fdc980140efbb8">terug</italic>) or were (<italic id="italic-142c8b0becc4816d3531587393c91d3f">weer</italic>) also used for the returnative sense and as we will see in the next section, the BP expressions that we discussed here are also used for the repetitive sense.</p>
      </sec>
      <sec id="heading-43e5f6324b750f8a251c51c070a54750">
        <title>2.6 Repetitive sense</title>
        <p id="paragraph-1c343dd893912b0c517821136a70bd2f">At the border of the counterdirectional domain we find the repetitive sense.</p>
        <p id="paragraph-a90cf75a1014bb02c676620f07a2c172">For BP, we find the exact same expressions as we saw in the last section, about the restitituve sense, namely <italic id="italic-c4ea2ebef5c2575a678431086a30008b">de novo</italic>, <italic id="italic-047e019b7d5eaee0c7a9a9542ad457e4">novamente</italic>, <italic id="italic-88831f1cb0fe788c3b24d8b5c0172d52">outra vez</italic>, <italic id="italic-824afc4fc09b2ef540b3728e510bc22f">mais uma vez</italic>, and also some <italic id="italic-91b91885b97c1f7bdff54b3f287f274d">re- </italic>forms lexically specialized.</p>
        <p id="paragraph-63ecccde94374bc144cd8684863a3f92">.</p>
        <p id="paragraph-40c9e8237517a265cda54105b0251d69">Não devemos nunca mais nos ver de novo / novamente / outra vez.</p>
        <p id="paragraph-51fafcf5079f300a3a3faef97f6e0644">We should never see each other again / another time.</p>
        <p id="paragraph-176e8bb862c70d124c68d5741b3f13c3"> .</p>
        <p id="paragraph-3b3f8282b7d6ac055720721ef72a53d5">As a first approximation, we can say that the difference between repetitive and restitutive sense in BP is not dependent on the adverbial chosen, but on the form of the verb. Consider again the verb <italic id="italic-9907fac59535208439e001ca3f778487">abrir </italic>(‘to open’), and the two sentences below:</p>
        <p id="paragraph-cd0d53948b11b7c14e559bf7bd54f0ee">(19) A porta abriu.</p>
        <p id="paragraph-e2acadc1aa3a8bfd112acdd1fea0fe9e">The door opened.</p>
        <p id="paragraph-545d457cfa03b77dc332e627f2785821">(20) A porta está aberta.</p>
        <p id="paragraph-bf7120dff56aaa87eb5354dde993708a">The door is opened</p>
        <p id="paragraph-57a630ae69e446dda64c5515861557e0"> .</p>
        <p id="paragraph-14968469b737989a7ff25acd02ca3e02">Sentence (20) means basically the same thing as sentence (19), but the difference is that (19) emphasizes the opening event and (20) the state of being open (<italic id="italic-b8cde7808c41a9b3c9a546d4fd1aff35">aberta </italic>is the feminine past participle form of the verb <italic id="italic-e5bd7d42c4da73711aeb5c245a4ad973">abrir</italic>). Both sentences can be combined with <italic id="italic-0fbbd34169a8002c68a05f181427c897">de novo</italic>/<italic id="italic-a64eb294bea7032b41a858486be15192">novamente </italic>‘again’, <italic id="italic-c6a391a0b3ca8f4036b0e0d33fee2756">outra vez </italic>‘another time’ and <italic id="italic-e0129b53e9f107d2d8d03cc0509230cb">mais uma vez </italic>‘once more’/‘one more time’:</p>
        <p id="paragraph-35cb32e63996dfe06799eacb67861baa">(19’) A porta abriu de novo <italic id="italic-5d3acaae0b38f7775e7e5706e918d01c">/ </italic>outra vez <italic id="italic-5f6aa94d967592aba5df66a0a9166b65">/ </italic>mais uma vez.</p>
        <p id="paragraph-f3cbc0dc7a5e8aa91cc99aa255f4de1f" />
        <p id="paragraph-d31749fd51dc45cb83be12670ef88f15">(20’) A porta está aberta de novo <italic id="italic-43b9d5ce77da4b2b6edbafbc47876213">/ </italic>outra vez <italic id="italic-666bd7b6643d3b30d3f49113989c1d3a">/ </italic>mais uma vez.</p>
        <p id="paragraph-1fbdb0b6d90d013f0c1b73d9ccb119a6" />
        <p id="paragraph-8083e7de76f470f9d9fb98097b94dfbb">Although both sentences are ambiguous between a restitutive and a repetitive sense, sentence (19’) is interpreted more often with the repetitive sense, and sentence in (20’) with the restitutive sense.</p>
        <p id="paragraph-b9603e66351e65466582e0c789305d57">In standard Dutch we find <italic id="italic-519f19c11956798ca2ad633f25c7c65e">weer</italic>, but in non-standard Belgian Dutch <italic id="italic-1d35680b99a19566b7301cd5012baa5c">terug </italic>is possible too. In addition, we also find in standard Dutch the possibility of using <italic id="italic-8c01b723e56ccd59962c93a17462951f">opnieuw </italic>‘anew’ or <italic id="italic-d84d73c6a967068774c6a33ca49b5ea1">nogeenkeer </italic>‘yet another time (lit. turn)’, specifically for the repetitive sense, as well. The latter adverbs are always stressed while <italic id="italic-0c7b01f9d71ae6ba2658722c9650d55a">weer </italic>and <italic id="italic-f8a12dbcb530e90f5dc184dd25059b76">terug </italic>can remain unstressed. This also affects the syntactic positions of these items, which we will not discuss here. Finally, the prefix <italic id="italic-5fff6a32a3a4aadee3f792da5e9ff530">her- </italic>also figures here.</p>
        <p id="paragraph-1fd5b4fd84cbf6cc196b213a2fa52f66">(21) </p>
        <p id="paragraph-e67e57616015e0e1f1e86e1e426648ba">a. Het voorstel is terug/weer/opnieuw/nog een keer afgewezen.</p>
        <p id="paragraph-655affc01a8ce7cd068e46f0ee86da0e">the proposal is terug/weer/opnieuw/nog een keer rejected</p>
        <p id="paragraph-22685f04437969f0eda9a2b0931b45cd">‘The proposals has been rejected again.’</p>
        <p id="paragraph-59e045dc377954ae464944d2a330ee58">b. Volgend jaar vindt het congres terug/weer/opnieuw/nog een keer in Gent plaats.</p>
        <p id="paragraph-09399f2c4c0b2e6c97a580bbdb3e1fa9">next year finds the congress terug/weer/opnieuw/nog een keer in Gent place</p>
        <p id="paragraph-f3861bd92547950746ccb4937a269a1a">‘Next year the congress will take place in Ghent again.’</p>
        <p id="paragraph-56f9023ccffe8e22b14c9601b753b2db">c. Ik heb het boek herlezen.</p>
        <p id="paragraph-31d5ad9bcf0d80f4870a3b5549509e90">I have the book her-read</p>
        <p id="paragraph-18ebb43bfbd502d2c10cf2b9030e4512">‘I reread the book.’</p>
        <p id="paragraph-aa464234755d3ef8086df7e531a48ef7"> .</p>
        <p id="paragraph-7cdd70be03afa4718a5d725cb92ecc47">In standard Dutch, <italic id="italic-c8bebeadf3d1e10457f0a2471a166bde">terug </italic>is only possible in the repetitive sense with the verbs <italic id="italic-2bf2cefdbcca8179d21c0f6762117fad">bellen </italic>‘phone’ and <italic id="italic-b9b3af70d610cb73a4ee3b236347a3c7">zien </italic>‘see’ (and so are <italic id="italic-34350d60153621111ce0ad4f0246c2c1">weer</italic>, <italic id="italic-5fd959ee08fabcd3092c18daf8efe968">opnieuw</italic>, <italic id="italic-9dd37dcb8627af02da69c549ef88752a">nogeenkeer</italic>):</p>
        <p id="paragraph-f699e114e1569d9b8d7ec025572a8b45"> .</p>
        <p id="paragraph-4c96d275f3a36f8c1b1e87dd48d5bbc6">(22) </p>
        <p id="paragraph-ddad303e68018f4c22d2ee5997bf07ce">a. Wij zullen elkaar nooit terug zien. we shall each-other never terugsee ‘We shall never see each other again.’</p>
        <p id="paragraph-19">b. Hij nam niet op dus ik zal hem terug bellen.</p>
        <p id="paragraph-379dbbc068f90fc35c33b3caea143c63">hetook not up so I will him terugcall</p>
        <p id="paragraph-21">‘He did not answer, so I will call him again’</p>
        <p id="paragraph-ee054735dee93c86daf45470147c0c99">.</p>
        <p id="paragraph-3e82f21c3cc3fb8042d382e2f007fe97">With the repetitive sense, the number of different terms used is relatively rich. We find five different forms in the Dutch language area and four different expressions in BP. The different terms might not be fully equivalent. In addition to differences in register or region, there might be differences in syntactic and prosodic behavior, and items might also have a collocational connection with particular verbs.</p>
      </sec>
    </sec>
    <sec id="heading-1e7f1e89881bd25d56de29371df0a604">
      <title>Conclusions</title>
      <p id="paragraph-2d8657d47a187cedd983b3051fad0189">Some of our conclusions are shown in the table and figures below; first, a simple tabular comparison between Dutch and Brazilian Portuguese, and then a schematic depicture of the distribution of the expressions over the map of senses in both languages. In Table 1 we summarize the expressions that we have found for the six senses. Parentheses enclose items that are archaic or regional.</p>
      <table-wrap id="table-figure-c0b1325d6f8e86524be09c8d664293d6">
        <label>Table 1</label>
        <caption>
          <title>Counterdirectional expressions in Brazilian Portuguese and Dutch</title>
          <p id="paragraph-6e12fbb1202acc15a64cdefff3110f27" />
        </caption>
        <table id="table-68a1af164ec924181098d12bca54a867">
          <tbody>
            <tr id="table-row-9cac68cb44e43815361c335bc374cf48">
              <th id="table-cell-cd0846c9b121e97ae73b19b7925189b4">SENSE</th>
              <th id="table-cell-910c5e0e65bcc04da2a2ba6a4da982c2">BrPORT<xref id="xref-12a026c41aa353d370a06d117812f0d2" ref-type="table-fn" rid="footnote-f5c66752767166dbca72b4f49e5117ba">?</xref></th>
              <th id="table-cell-ab9f0371a9656147793c3eb8a47dcd21">DUTCH</th>
            </tr>
            <tr id="table-row-41a9cb10a0437ea3e67491f6d079ebd1">
              <td id="table-cell-44baca4772e1a951c3ab1c48ebeb2e7f">REAWARD</td>
              <td id="table-cell-e7ad40a00d0e363e53cd9c1476628764">re-, de costas, para trás</td>
              <td id="table-cell-9bd83050668213a85d62f25a04448d68">achteruit, terug</td>
            </tr>
            <tr id="table-row-d29ccf01c826315ea1e0ca0c1412320c">
              <td id="table-cell-004e7ddad826af431ddc961a3dd04829">RETROGRADE</td>
              <td id="table-cell-fb8fdd3b661c2830efa9b658b737ed34">re-, de trás para frente, ao contrário, para trás</td>
              <td id="table-cell-884998bcf038916bf1107858210cc414">achteruit, terug</td>
            </tr>
            <tr id="table-row-a97cb36d39fd584e9dcbc9d3728ec25a">
              <td id="table-cell-615c877541d73f50012be56e5bc23161">RETURNATIVE</td>
              <td id="table-cell-4731095ee23a828fe1d8545f4888972d">re-, de volta, voltar, tornar</td>
              <td id="table-cell-0af2d66e30c2b0d50ed90daad1fe8648">terug, (weer)</td>
            </tr>
            <tr id="table-row-c7cc16e761d038ab3fbcb621e9041315">
              <td id="table-cell-fa3bd51b8a3df2235efa114fd4110912">RESPONSIVE</td>
              <td id="table-cell-073e2040a5d0a64d9f380a42b1037bc2">re-, de volta, devolver, voltar</td>
              <td id="table-cell-bba83839e9115d4f521a7a39f71eda70">terug</td>
            </tr>
            <tr id="table-row-419c88f95a81584d71be6b894f46a4af">
              <td id="table-cell-bd545d19e4da4a2861eedbfce8f5a7f7">RESTITUTIVE</td>
              <td id="table-cell-22764ee51f5668807c13b44739ae7ddf">re-, de novo, novamente, outra vez, mais uma vez</td>
              <td id="table-cell-4bf10fc3a0d71a06feda59b744b27b62">weer, her-, (terug)</td>
            </tr>
            <tr id="table-row-b58556c6d180f2783fd5c131e4f97cd9">
              <td id="table-cell-d5cf88988cb890ea4d3337eaea056644">REPETITIVE</td>
              <td id="table-cell-262de4b7fccaf8c16fbeed9820c52d48">re-, de novo, novamente, outra vez, mais uma vez</td>
              <td id="table-cell-6e4badf27cffd2c826467e113753182c">weer, her-, opnieuw, nog een keer, (terug)</td>
            </tr>
          </tbody>
        </table>
        <table-wrap-foot>
          <fn-group>
            <fn id="footnote-f5c66752767166dbca72b4f49e5117ba">
              <label>*</label>
              <p id="paragraph-d7872d08f6e7956d3e2cf9f61a0b8d8a">For BP, forms with the prefix re- can be found for all the senses with different levels of compositionality and productivity, and are usually more formal due to their lexical specialization. For these reasons, we leave them aside in the next figures.</p>
            </fn>
          </fn-group>
        </table-wrap-foot>
      </table-wrap>
      <p id="paragraph-942aada770d38b1bfa72ae274a2aaf11">The (standard, modern) Dutch situation is represented in Figure 7. In Belgian Dutch <italic id="italic-e15723554d80ffd954b5a9f90e2c8fb8">terug </italic>covers the whole map and in more archaic Dutch, <italic id="italic-36bc1eb0b7691094af3eb04bc9b23964">weer </italic>also covers the returnative sense.</p>
      <fig id="figure-panel-3e9f3dc55f6be5a7567e190d9e7d5b50">
        <label>Figure 6</label>
        <caption>
          <title>Counterdirectionality in Dutch</title>
          <p id="paragraph-6b88da8508bdfd3aa439d1d5cbf21ed4" />
        </caption>
        <graphic id="graphic-f9518326510f1aa990a070e02c11d0f6" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_12-07-39.png" />
      </fig>
      <p id="paragraph-b5c4ba468ca55b22dd937c4ae3499818">The Brazilian Portuguese situation is found in Figure 8. We do not represent the distribution of the prefix <italic id="italic-c7993adaf841578788f232f6d638d019">re-</italic>, that covers the whole map.</p>
      <fig id="figure-panel-52a8f6d12eb03967631ae033a8a449e8">
        <label>Figure 7</label>
        <caption>
          <title>Counterdirectionality in Brazilian Portuguese</title>
          <p id="paragraph-2c428de5e915c7e5eaa77a6a0cf6dba1" />
        </caption>
        <graphic id="graphic-01ab710cdc2bb6550d39193aa9178827" mimetype="image" mime-subtype="png" xlink:href="image_2021-07-17_12-11-26.png" />
      </fig>
      <p id="paragraph-a0cef050620c9278081f18c9eb72f214">As we can see, these two languages divide up the space of meanings in quite different ways. While BP has kept the senses we explored in this paper in three clearly separated categories (namely, (i) rearward and retrograde, (ii) returnative and responsive, and (iii) restitutive and repetitive), Dutch is a bit more complicated, because it shows overlapping categories. Nevertheless, each of the lexical regions in both languages is constrained by covering a contiguous set of senses, an important property for semantic maps (Haspelmath 2003).</p>
      <p id="paragraph-0004362d693d3057f3fd49b3c9b70905">Interestingly, Brazilian Portuguese and Dutch have partially followed different trajectories in how they express meanings in the domain of counterdirectionality. There is not one unique path of historical development leading from the most concrete ‘back’ notion to more abstract notions, but counterdirectional expressions are recruited from different semantic sources. In BP the idea of rotation is very important, as attested by the forms based on <italic id="italic-3d71f07f322a0ff76b633b23f0b7a1c7">voltar</italic>, and this is the base for the returnative and responsive senses.<xref id="xref-2efa2aec444e67f6a41cd01f76718494" ref-type="fn" rid="footnote-e9711e2602f2fb0cfc702676dcc112e1">10</xref> In Dutch, however, ‘back’ is the source for these senses, while the role of ‘back’ is more limited in BP. There are also strong similarities in the way ‘new’ (<italic id="italic-ca5b361b4ce37a5efc1ab352bb2a45d4">novo</italic>, <italic id="italic-0981329433cfbf41f5326bf461453a0c">nieuw </italic>is used for the repetitive sense. An important finding is that BP has extended this use also to the restitutive sense.</p>
      <p id="paragraph-ee7b1824d6d76707add6e7b07d584755">It would be interesting to find out what types of expressions other languages recruit to express different senses in the counterdirectional domain, in addition to English, BP, and Dutch. There is some typological literature that shows how a verb for ‘return’ can develop into a returnative or repetitive marker in many languages (Lichtenberk 1991, Moyse-Faurie 2012). As far as we know, the reverse direction of inquiry has not been explored: start with the different senses within the counterdirectional domain, to find out how they are expressed in different languages (although Wälchli’s 2006 exploration of parallel corpora based on the gospel of Mark could be a good starting point here).</p>
      <p id="paragraph-f469f3d4d7a49472ef7dfb819c0f36fb">Although our theoretical basis in this paper is informal and the our treatment of the data restricted in certain ways, we believe to have shown that there is a coherent (sub)domain of senses that can be studied across languages with interesting results for typology and lexical semantics, and potentially also for grammaticalization theory and morphosyntax. More work is needed to understand the way counterdirectional elements are specialized for particular verbs or senses in ways that go beyond the six senses identified here. Also, the differences between BP and Dutch need to be put in a broader historical and typological context, taking into account other Germanic and Romance languages.</p>
    </sec>
  </body>
  <back>
    <fn-group>
      <fn id="footnote-ba3a5a595694743ca546c5640ab3dbc3">
        <label>1</label>
        <p id="paragraph-68859b0e1b34817787a1e0052dfd06bd">We ignore an additional metaphoric sense that involves development towards a worse state, attested for English backward, as in The party went backward, and also for Brazilian Portuguese and Dutch counterparts.</p>
      </fn>
      <fn id="footnote-b877bcaa5aabc7c10c8dd7227b7700cf">
        <label>2</label>
        <p id="paragraph-c569fc4a4a377f6a01db4ab89cb6716f">Because motion and action share a general notion of directionality and because motion is more ‘basic’ than action, there is a fundamental conceptual metaphor ACTION IS MOTION across languages in the Conceptual Metaphor Theory (Lakoff &amp; Johnson 1980). Within that theory the responsive sense of <italic id="italic-aad5edc48fcd3bd148cd8070cc0df9b0">back </italic>is an instance of that metaphor. For us, the deeper explanation of that metaphor is the shared directionality of motion and action.</p>
      </fn>
      <fn id="footnote-5b3ba0fe95173e1f9c9c6c9f3a0d92d2">
        <label>3</label>
        <p id="paragraph-7e4d9940c5869ef6c6a8d57f9e546757">This is the basis of the conceptual metaphor CHANGE IS MOTION in the Conceptual Metaphor Theory (Lakoff &amp; Johnson 1980), with restitutive <italic id="italic-f3eb582ae258933f5f92ccbc9cbe53bc">back </italic>as an example. For us, the returnative and restitutive concepts are related to each other, not because there is a conceptual metaphor in the language that relates these concepts, but because they are related by shared directional structure.</p>
      </fn>
      <fn id="footnote-799c528c9f5c041c0439dd6e76e4c0a3">
        <label>4</label>
        <p id="paragraph-784330d325fdbd6e614eccf192d99b43">In BP the preposition <italic id="italic-66b3e26d3af2e66ad6ea66930b160729">para </italic>can also be written <italic id="italic-14828e7f5912e24037db7c61aa26a391">pra</italic>; this is just a matter of sociolinguistic variation.</p>
      </fn>
      <fn id="footnote-e8bce4dac02d3d44e931918744ada1e9">
        <label>5</label>
        <p id="paragraph-3fa9a7434951259536082ac65ecf10ec">According to etymological dictionaries, the BP verb <italic id="italic-8d184df072063ae6358250454f07a0e9">recuar </italic>comes from the French verb <italic id="italic-c3602be1dfbd499924f271b2e0e21e2d">reculer </italic>(which means to retreat, to recoil, to back away), and <italic id="italic-b291183acc805829a74fe6cb599f8dc3">reculer </italic>has probably derived from the Latin form <italic id="italic-7239ba804742952d540c3895a74aa015">culus</italic>, <italic id="italic-9b3e769682b07cc1d94f943ad285d6e7">i </italic>which indicates the part of the body used to sit upon, on the back of the body, which one cannot see, i.e. the buttocks, through a form like *<italic id="italic-58482a23dc1a56f4bd0c2b9ce8ee99a3">reculare</italic>. We use “probably” because we couldn’t find attested instances of these forms.</p>
      </fn>
      <fn id="footnote-b0c1e335e24c911deff3abb9eb6d0230">
        <label>6</label>
        <p id="paragraph-c2478db6d33d7e3da4fe472eae67af53">Grammars of BP traditionally recognize and treat <italic id="italic-815d6f077118bca7bfe089df78f9fc2b">re- </italic>as a “repetitive morpheme”, but things are much more complicated. In fact, this morpheme can be used to indicate repetition (as we’ll see in section 3.6) but also other nuances of meaning related to movement. It is not our aim in this paper to discuss the semantics of this prefix, but we’ll consider, for simplification, that its compositional meaning is repetition and all other nuances of meaning are achieved via different mechanisms.</p>
      </fn>
      <fn id="footnote-10d15baec9113ac81544071f6e239f87">
        <label>7</label>
        <p id="paragraph-665724d1fe03febe8d99df8d258e31c1">
          <ext-link id="external-link-7909ad5504af2d1b6e8f57f5f93058ce">http://noticias.uol.com.br/tabloide/ultimas-noticias/tabloideanas/2014/06/06/indiano-anda-de-costas-por-25-anos-e-esquece-como-e-andar-para-frente.htm</ext-link>
        </p>
      </fn>
      <fn id="footnote-2a4d93373603971d3a2728289d748f63">
        <label>8</label>
        <p id="paragraph-dfab711046367b1d1ec321281ba02c33">In this case, the re- form, <italic id="italic-2ccfb372583b4c48bcee06e7be4e6b87">reverter</italic>, is more formal, and not strictly compositional, since the verb <italic id="italic-80f90e508537e55aaabe8a00a8add44f">verter </italic>means to pour (water/wine/liquid).</p>
      </fn>
      <fn id="footnote-5de877820a4606f36e8c9241b541d6b6">
        <label>9</label>
        <p id="paragraph-034868f56c63d1afff94fe179c92d343">The form *<italic id="italic-134b3397bbb7ca9d05d689bcbc2552b6">voltare </italic>probably did not exist in Classic Latin, but must have appeared in Vulgar Latin originating from <italic id="italic-48cd98d52aea0fcd34d8080029cad848">voltum </italic>which is the supine of the verb <italic id="italic-b6459ac28adb22daedaffe3afda8a109">volvere</italic>, which means to roll, to revolve.</p>
      </fn>
      <fn id="footnote-e9711e2602f2fb0cfc702676dcc112e1">
        <label>10</label>
        <p id="paragraph-5e990ae43fcd8d293a8881cec5aa16bc">In Dutch, the notion of rotation plays a small role, in the etymology of the noun <italic id="italic-8183561e73743f966221203d1a43a006">keer </italic>‘turn’ in <italic id="italic-6620d9e8e21dcfc61d7de1a11ec41d64">nogeenkeer</italic>, literally ‘yet another turn’ for the repetitive sense. We have not explored the historical trajectory of that expression.</p>
      </fn>
    </fn-group>
    <ref-list>
      <ref id="journal-article-ref-6b7376870179586d159bbba6651589f0">
        <element-citation publication-type="journal">
          <fpage>3</fpage>
          <issue>1</issue>
          <lpage>51</lpage>
          <volume>22</volume>
          <year>2005</year>
          <person-group person-group-type="author">
            <name>
              <surname>BECK</surname>
              <given-names>S</given-names>
            </name>
          </person-group>
          <source>Journal of Semantics</source>
          <article-title><bold id="bold-1">There and back again</bold>: A semantic analysis</article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-c7a70099c9a91ad9b529070e2aa4371c">
        <element-citation publication-type="journal">
          <fpage>193</fpage>
          <issue>3</issue>
          <lpage>214</lpage>
          <volume>12</volume>
          <year>2009</year>
          <person-group person-group-type="author">
            <name>
              <surname>BECK</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Berezovskaya</surname>
              <given-names>P</given-names>
            </name>
            <name>
              <surname>Pflugfelder</surname>
              <given-names>K</given-names>
            </name>
          </person-group>
          <source>Syntax</source>
          <article-title><bold id="bold-1dda3d960d8dbed2afbe55417b3af308">The Use of Again </bold>in 19th‐Century English versus Present‐Day English</article-title>
        </element-citation>
      </ref>
      <ref id="book-ref-589e0a7b565d884f943ac6128c5f4bdd">
        <element-citation publication-type="book">
          <publisher-loc>Dordrecht</publisher-loc>
          <publisher-name>Reidel</publisher-name>
          <year>1979</year>
          <person-group person-group-type="author">
            <name>
              <surname>DOWTY</surname>
              <given-names>D. R</given-names>
            </name>
          </person-group>
          <source><bold id="bold-699e3b596fa5548c134bd28ddd8d8669">Word meaning and Montague grammar</bold><italic id="italic-c355380af29e04a7a076dd30555e63a7">: </italic>The semantics of verbs and times in generative semantics and in Montague’s <italic id="italic-42568f77c4e3e41bba206d3b16eebff2">PTQ</italic></source>
        </element-citation>
      </ref>
      <ref id="chapter-ref-56aa0ec810b42d432b317eaf5c6d0555">
        <element-citation publication-type="chapter">
          <fpage>101</fpage>
          <lpage>130</lpage>
          <publisher-loc>Berlin</publisher-loc>
          <publisher-name>Akademi Verlag</publisher-name>
          <year>2001</year>
          <person-group person-group-type="author">
            <name>
              <surname>FABRICIUS-HANSEN</surname>
              <given-names>C</given-names>
            </name>
          </person-group>
          <person-group person-group-type="editor">
            <name>
              <surname>Fery</surname>
              <given-names>C</given-names>
            </name>
            <name>
              <surname>Sternefeld</surname>
              <given-names>W</given-names>
            </name>
          </person-group>
          <source>Audiatur Vox Sapientiae. A Festschrift for Arnim von Stechow</source>
          <chapter-title>
            <bold id="bold-fc152785bbdbc4a87fa942847e0124e1">Wi(e)der and again(st)</bold>
          </chapter-title>
        </element-citation>
      </ref>
      <ref id="chapter-ref-e746f57c04ddf35f0d468ae0e593b1cc">
        <element-citation publication-type="chapter">
          <fpage>211</fpage>
          <lpage>242</lpage>
          <publisher-loc>Mahwah, NJ</publisher-loc>
          <publisher-name>Erlbaum</publisher-name>
          <volume>2 </volume>
          <year>2003</year>
          <person-group person-group-type="author">
            <name>
              <surname>HASPELMATH</surname>
              <given-names>M</given-names>
            </name>
          </person-group>
          <person-group person-group-type="editor">
            <name>
              <surname>Tomasello</surname>
              <given-names>M</given-names>
            </name>
          </person-group>
          <source>The New Psychology of Language</source>
          <chapter-title><bold id="bold-9890cc19ab367222697e222f4f5bdbfe">The geometry of grammatical meaning</bold>: Semantic maps and cross-linguistic comparison</chapter-title>
        </element-citation>
      </ref>
      <ref id="book-ref-e05a9cfc734aa891d94d8a1e443561ca">
        <element-citation publication-type="book">
          <publisher-loc>Cambridge</publisher-loc>
          <publisher-name>Cambridge university Press</publisher-name>
          <year>2002</year>
          <person-group person-group-type="author">
            <name>
              <surname>HEINE</surname>
              <given-names>B</given-names>
            </name>
            <name>
              <surname>Kuteva</surname>
              <given-names>T</given-names>
            </name>
          </person-group>
          <source>
            <bold id="bold-95741dfdee2b219ddeec4ca000adae72">World Lexicon of Grammaticalization</bold>
          </source>
        </element-citation>
      </ref>
      <ref id="book-ref-87a25aff3d1a158f887a76cfdb55c684">
        <element-citation publication-type="book">
          <publisher-loc>Chicago</publisher-loc>
          <publisher-name>Chicago university Press</publisher-name>
          <year>1980</year>
          <person-group person-group-type="author">
            <name>
              <surname>LAKOFF</surname>
              <given-names>G</given-names>
            </name>
            <name>
              <surname>Johnson</surname>
              <given-names>M</given-names>
            </name>
          </person-group>
          <source>
            <bold id="bold-88b45c0c09b8dba2937d9c435a298ce4">Metaphors We Live By</bold>
          </source>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-37213626e7e5ffb8b6e10eb2380d64b0">
        <element-citation publication-type="journal">
          <fpage>475</fpage>
          <lpage>509</lpage>
          <volume>67</volume>
          <year>1991</year>
          <person-group person-group-type="author">
            <name>
              <surname>LICHTENBERK</surname>
              <given-names>F</given-names>
            </name>
          </person-group>
          <source>Language</source>
          <article-title>
            <bold id="bold-5c7ecaef14ba63f190cf7f3a2453ebdc">Semantic change and heterosemy in grammaticalization</bold>
          </article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-da2e361b918f1c2f1a888a8ea7ded987">
        <element-citation publication-type="journal">
          <fpage>11</fpage>
          <lpage>31</lpage>
          <volume>6</volume>
          <year>1995</year>
          <person-group person-group-type="author">
            <name>
              <surname>KEITH</surname>
              <given-names>Allan</given-names>
            </name>
          </person-group>
          <source>Cognitive Linguistics</source>
          <article-title>
            <bold id="bold-97c23cb7fe76a778a9220c7bf19517f9">The anthropocentricity of the English word(s) <italic id="italic-48c24798cdf13556599105bc3c31eeb3">back</italic></bold>
          </article-title>
        </element-citation>
      </ref>
      <ref id="conference-paper-ref-1cdb2b7bf4e640da8a2e6da04e66d691">
        <element-citation publication-type="confproc">
          <conf-name>P-workshop</conf-name>
          <conf-loc>University of Stuttgart</conf-loc>
          <year>2011</year>
          <person-group person-group-type="author">
            <name>
              <surname>MCINTYRE</surname>
              <given-names>A</given-names>
            </name>
          </person-group>
          <source>Paper</source>
          <article-title>
            <bold id="bold-2958589c540a524b0cf8fc8711f621fd">Particle incorporation in English double object constructions, nominalisations and else where</bold>
          </article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-06fb5a96d7c6e9bc7fe352c720d4365a">
        <element-citation publication-type="journal">
          <fpage>234</fpage>
          <lpage>260</lpage>
          <volume>51</volume>
          <year>2012</year>
          <person-group person-group-type="author">
            <name>
              <surname>MOYSE-FAURIE</surname>
              <given-names>C</given-names>
            </name>
          </person-group>
          <source>Oceanic Linguistics</source>
          <article-title>T<bold id="bold-1311fdae035c3f8c31469234cfbb0996">he concept ‘return’ as a source of different developments in Oceanic languages</bold></article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-bf27ae436585b6de168b72faab067109">
        <element-citation publication-type="journal">
          <fpage>373</fpage>
          <issue>3</issue>
          <lpage>424</lpage>
          <volume>32</volume>
          <year>2014</year>
          <person-group person-group-type="author">
            <name>
              <surname>PEDERSEN</surname>
              <given-names>W. A</given-names>
            </name>
          </person-group>
          <source>Journal of Semantics</source>
          <article-title>
            <bold id="bold-f312edb39af7ca576ced88739853f3d3">A scalar analysis of again-ambiguities</bold>
          </article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-7f284dd2b12d7e4fa4e251d9cc5b3459">
        <element-citation publication-type="journal">
          <fpage>87</fpage>
          <issue>2</issue>
          <lpage>138</lpage>
          <volume>13</volume>
          <year>1996</year>
          <person-group person-group-type="author">
            <name>
              <surname>VON STECHOW</surname>
              <given-names>A</given-names>
            </name>
          </person-group>
          <source>Journal of semantics</source>
          <article-title><bold id="bold-21970df534d4f7847ed0695a3bd8e7a6">The different readings of wieder ‘again’: </bold>A structural account</article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-56c201481227d61866a9bd83f17adafd">
        <element-citation publication-type="journal">
          <fpage>69</fpage>
          <lpage>113</lpage>
          <volume>30</volume>
          <year>2006</year>
          <person-group person-group-type="author">
            <name>
              <surname>WÄLCHLI</surname>
              <given-names>B</given-names>
            </name>
          </person-group>
          <source>Studies in Language</source>
          <article-title>
            <bold id="bold-29d28fe0de63488174fb3c30c21ff337">Typology of heavy and light again or the eternal return of the same</bold>
          </article-title>
        </element-citation>
      </ref>
      <ref id="journal-article-ref-742bdb0b915de86684ef16e238c5cf5e">
        <element-citation publication-type="journal">
          <fpage>377</fpage>
          <issue>1</issue>
          <lpage>395</lpage>
          <volume>8</volume>
          <year>2010</year>
          <person-group person-group-type="author">
            <name>
              <surname>ZWARTS</surname>
              <given-names>J</given-names>
            </name>
          </person-group>
          <source>Linguistic Discovery</source>
          <article-title><bold id="bold-b0dbe3376938c9f1ba4acec20e880cd2">Semantic map geometry: </bold>Two approaches</article-title>
        </element-citation>
      </ref>
    </ref-list>
  </back>
</article>