Constraints In Discourse 2 Peter Khnlein Ed Anton Benz Ed

Constraints In Discourse 2 Peter Khnlein Ed
Anton Benz Ed download
https://ebookbell.com/product/constraints-in-discourse-2-peter-
khnlein-ed-anton-benz-ed-1892588
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Constraints In Discourse 2 Peter Khnlein Ed Anton Benz Ed
https://ebookbell.com/product/constraints-in-discourse-2-peter-
khnlein-ed-anton-benz-ed-4715074
Constraints In Discourse 172th Edition Anton Benz Ed Peter Khnlein Ed
https://ebookbell.com/product/constraints-in-discourse-172th-edition-
anton-benz-ed-peter-khnlein-ed-2010442
Constraints In Discourse 3 Representing And Inferring Discourse
Structure Anton Benz Ed
https://ebookbell.com/product/constraints-in-discourse-3-representing-
and-inferring-discourse-structure-anton-benz-ed-4688972
Constraints In Computational Logics Theory And Applications
International Summer School Ccl 99 Gifsuryvette France September 58
1999 Revised Lectures 1st Edition Jeanpierre Jouannaud
https://ebookbell.com/product/constraints-in-computational-logics-
theory-and-applications-international-summer-school-
ccl-99-gifsuryvette-france-september-58-1999-revised-lectures-1st-
edition-jeanpierre-jouannaud-4198638

Constraints In Phonological Acquisition 1st Edition Rene Kager
https://ebookbell.com/product/constraints-in-phonological-
acquisition-1st-edition-rene-kager-1715176
Functional Constraints In Grammar On The Unergativeunaccusative
Distinction Susumu Kuno
https://ebookbell.com/product/functional-constraints-in-grammar-on-
the-unergativeunaccusative-distinction-susumu-kuno-1715714
Accounting For Constraints In Delay Systems Giorgio Valmorbida
https://ebookbell.com/product/accounting-for-constraints-in-delay-
systems-giorgio-valmorbida-47165958
Choices And Constraints In Family Life 2nd Edition Maureen Baker
https://ebookbell.com/product/choices-and-constraints-in-family-
life-2nd-edition-maureen-baker-34720864
Moral Animals Ideals And Constraints In Moral Theory Catherine Wilson
https://ebookbell.com/product/moral-animals-ideals-and-constraints-in-
moral-theory-catherine-wilson-1470702

Volume 194
Constraints in Discourse 2
Edited by Peter Kühnlein,Anton Benz and Candace L. Sidner
Founding Editors
Jacob L. Mey
University of Southern
Denmark
Herman Parret
Belgian National Science
Foundation, Universities of
Louvain and Antwerp
Jef Verschueren
Belgian National Science
Foundation,
University of Antwerp
Pragmatics & Beyond New Series is a continuation of Pragmatics & Beyond and
its Companion Series. The New Series offers a selection of high quality work
covering the full richness of Pragmatics as an interdisciplinary field, within
language sciences.
Pragmatics & Beyond New Series (P&BNS)
Editorial Board
Robyn Carston
University College London
Thorstein Fretheim
University of Trondheim
John C. Heritage
University of California at Los
Angeles
Susan C. Herring
Indiana University
Masako K. Hiraga
St. Paul’s (Rikkyo) University
Sachiko Ide
Japan Women’s University
Kuniyoshi Kataoka
Aichi University
Miriam A. Locher
Universität Basel
Sophia S.A. Marmaridou
University of Athens
Srikant Sarangi
Cardiff University
Marina Sbisà
University of Trieste
Deborah Schiffrin
Georgetown University
Paul Osamu Takahara
Kobe City University of
Foreign Studies
Sandra A. Thompson
University of California at
Santa Barbara
Teun A. van Dijk
Universitat Pompeu Fabra,
Barcelona
Yunxia Zhu
The University of Queensland
Editor
Anita Fetzer
University of Würzburg
Associate Editor
Andreas H. Jucker
University of Zurich

Constraints in Discourse 2
Edited by
Peter Kühnlein
function2form
Anton Benz
Centre for General Linguistics, Berlin
Candace L. Sidner
Worcester Polytechnic Institute
John Benjamins Publishing Company
Amsterdamâ•›/â•›Philadelphia

Library of Congress Cataloging-in-Publication Data
Constraints in discourse 2 / edited by Peter Kühnlein,Anton Benz and Candace L. Sidner.
p. cm. (Pragmatics & Beyond New Series, issn 0922-842X ; v. 194)
Includes bibliographical references and index.
1. Discourse analysis.2. Constraints (Linguistics) I.Kühnlein,Peter.II.Benz,Anton,1965-
III. Sidner, C. L. IV. Title: Constraints in discourse two.
P302.28.C67 2010
401’.41--dc22 2009047143
isbn 978 90 272 5438 2 (Hb ; alk. paper)
isbn 978 90 272 8854 7 (Eb)
© 2010 – John Benjamins B.V.
No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any
other means, without written permission from the publisher.
John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands
John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
The paper used in this publication meets the minimum requirements of
American National Standard for Information Sciences – Permanence of
Paper for Printed Library Materials, ansi z39.48-1984.
8
TM

Table of contents
Rhetorical structure: An introduction 1
Peter Kühnlein
Clause-internal coherence 15
Jerry R. Hobbs
Optimal interpretation for rhetorical relations 35
Henk Zeevat
Modelling discourse relations by topics and implicatures:
The elaboration default 61
Ekatarina Jasinskaja
The role of logical and generic document structure in relational
discourse analysis 81
Maja Bärenfänger, Harald Lüngen, Mirco Hilbert Henning Lobin
Obligatory presupposition in discourse 105
Pascal Amsili Claire Beyssade
Conventionalized speech act formulae: From corpus findings
to formalization 125
Ann Copestake Marina Terkourafi
Constraints on metalinguistic anaphora 141
Philippe De Brabanter
Appositive Relative Clauses and their prosodic realization in spoken
discourse: A corpus study of phonetic aspects in British English 163
Cyril Auran Rudy Loock
Index 179

Rhetorical structure
An introduction
Peter Kühnlein
0.1â•… General remarks
Texts, and in general types of discourse, vary along a multitude of dimensions.
Discourse can be spoken or written, monological or an exchange between a number
of participants, it can be employed to inform, persuade (and serve many more
or even mixed functions), it can take place in various settings and be arbitrarily
extensive. However, some characteristics are shared by all kinds of texts.
One of those shared properties is that text, and discourse in general, is struc-
tured, and they in turn are so in a multitude of ways: classical written text as
the present, e.g., typically has logical and graphical structuring into paragraphs,
sections, chapters etc. depending on the type of text or the genre, but it is by
nature monological. Spoken dialogue, at the other end of the spectrum, inter alia
is characterized by assignment of and changes in roles participants assume in the
exchange, stretches of overlapping speech, repairs and many more phenomena
that are not regularly observed in written text (notwithstanding chats on the internet
and the like) and which give rise to completely different types of structure. All of
these forms of communication fall under the common denominator discourse; we
will keep using this cover term here to refer to them.
The different kinds of structures in discourse have been object toresearch for
a considerable time. One type of structure has been of special interests for researchers
working in more formal paradigms and has been hotly discussed ever since: it is
what is called the rhetorical or coherence structure. Rhetorical structure is built
by applying rhetorical relations recursively to elementary units of discourse. This
kind of structure is to be distinguished from, e.g., cohesive structure that comes to
existence by means of, e.g., coreference in various forms.
The present collection comprises papers that give a wide variety of perspec-
tives on the constraints governing discourse structure, and primarily rhetorical
structure: various ways of thinking of constitutive units of discourse along with a
variety of conceptions of rhetorical relations are presented, and the issue of which
kind of structure is right for the description of the rhetorical make-up of discourse
is tackled from different points of view.

 Peter Kühnlein
Accordingly, this introductory chapter is intended to provide the necessary
background to understand the discussions by sketching as briefly as possible the
state of the art; the reader is referred to the individual chapters in the volume
where appropriate.
As the previous volume, Constraints in Discourse (Benz Kühnlein,
2008), the present one is the result of selecting and compiling papers that are
extended versions of presentations at a workshop in the series “Constraints
in Discourse.” The second of these workshops was held in Maynooth, Ireland,
and organized by Candace Sidner (chair), Anton Benz, John Harpur and Peter
Kühnlein. All the authors who contribute to the present volume submitted
their re-worked and substantially extended papers to a peer reviewing process,
where each author had to review two other authors’ papers. In addition, John
Benjamins conducted an own reviewing process before agreeing to publish the
collection. This two-stage reviewing process is intended to secure high quality
of the contributions.
0.2â•… Elementary units
Just as in any formal description of structures, one basic step in describing
rhetorical structure of a discourse is to identify the elementary units. Due to
the multitude of dimensions along which discourse can vary and due to differ-
ences in theoretical assumptions, there is no consensus on what to count as an
elementary unit.
Exemplarily, there is a divide between proposals for different domains: a
proposal for spoken discourse can refer to intonational features as an important
criterion for segment status, whereas a proposal made specifically for written dis-
course can’t. On the other hand, a proposal set up for written discourse can make
reference to punctuation and syntactic units, whereas the first is absent and the
the latter are not reliably correct in spoken discourse.
Research in prosodic features of discourse and its segments reaches far back:
Butterworth (1975) reports that speech rate changes during discourse segments,
being higher at the end of a segment than at the beginning. Chafe (1980) observes
that pause lengths are varying at segment boundaries too. Much corpus based
and computer linguistic research in this area was conducted by Julia Hirschberg
with various collaborators, e.g., Hirschberg and Pierrehumbert (1986), Grosz and
Hirschberg (1992), Hirschberg and Pierrehumbert (1992), or Hirschberg and
Nakatani (1996).
One of the most detailed empirical inquiries into the relation between
discourse segmentation and prosody is given in (Hirschberg Nakatani, 1996),

Rhetorical structure 
and the methodology employed there deserves a little closer description. The
authors set up a corpus of directives, where subjects had to give route descriptions
of varying complexity through Boston. The first series of descriptions was given
spontanously by the subjects, recorded and then transcribed. In a second series,
the same subjects read the corrected (i.e., freed of false starts etc.) transcripts of
their route descriptions, and again the speech was recorded and transcribed. The
data obtained from one of the speakers were prosodically transcribed using the
ToBI standard, split into intermediate phrases, pause lengths were measured
and fundamental frequencies (F0) and energy (RMS) calculated.
Then two groups of annotators marked up the texts with segment boundaries:
one group was given the transcripts only, the other group was given transcripts
plus recorded speech. The theory that served as background to segmenting the
texts was that of Grosz and Sidner (1986). Their account is potentially indepen-
dent from domain, i.e., applicable to both spoken and written discourse; Grosz and
Sidner claim that discourse structure actually consists of three distinct, but interacting
levels. The most central of these levels is the intentional one: for every coherent
discourse, that is the claim, one can identify an overarching discourse purpose the
initiating participant seeks to pursue. The segments of discourse according to this
theory correspond to sub-purposes, the so-called discourse segment purposes.
Elementary units in this theory correspond to single purposes. The two other
levels, attention and linguistic realization, concern which objects are in the
center of discourse and how the discourse is actually realized using cue-phrases
and special markers.
The results obtained by (Hirschberg Nakatani, 1996) in their study on spoken
discourse confirm previous findings and reveal much more detail than, e.g., the
work by Butterworth (1975); Chafe (1980) reports: both F0 and energy are higher
at the beginnings of discourse segments than at their ends. Speech rate on the
other hand increases towards the end of a segment, and pause lengths during a
segment are shorter than before a segment beginning and after a segment end. So
segment boundaries as judged according to purposes indeed seem to be correlated
with measurable changes in the speech signal.
These, and similar, findings seem to indicate good mutual support between
the intention-based theory of discourse structure developed by Grosz and Sidner
(1986) and the claim that discourse segment boundaries are marked intonationally
in spoken discourse.
Considerably more work than on spoken discourse has traditionally been
devoted to written text than to spoken discourse phenomena. The pioneering
work dates back to the 80s, and the cited work by Grosz and Sidner (1986) is
among that. An account of discourse structure that had comparable impact at
that time was developed by Mann and Thompson (1987a, 1988b). This account,

 Peter Kühnlein
known as Rhetorical Structure Theory (rst) was explicitly developed as a means
to capture analysts’ judgements about writers’ intentions while composing texts.
Thus, rst is devoted to the analysis of written text, but the analysis is not primar-
ily guided by linguistic surface structure: according to its founders, it is rather
“pre-realizational” in that it aims to describe the function of (the interplay of)
constituents in abstraction from linguistic realization.
Mann and Thompson (1987a, 1988b) claim that the base case for linguis-
tic expressions conveying intentions are clauses of certain types: main clauses,
non-restrictive relative clauses are of the right variety, whereas, e.g., restrictive
relative clauses and complement clauses (e.g., in subject or object position in a
matrix clause) are not counted as minimal units. It turned out in the develop-
ment of rst since its inception that narrowing down the type of constructions that
express writers’ intentions to clauses poses problems in multi-lingual applications:
what is expressed in a clause in one language might more suitably be expressed
in a different construction in another language. This case was made especially
by Rösner and Stede (1992) and Carlson and Marc (2001) who consequently
proposed extensions of the set of minimal units. The motivation for the inclusion
of certain constructions (or the exclusion of others) is not always readily understand-
able.So,researcherscomprisingCarlsonandMarcu(2001)andLüngenetal.(2006)
(cf. also paper 0.5) working in the rst paradigm, but likewise Wolf and Gibson
(2005,2006),optforincludingcertainpre-posedppslike“OnMonday,”inthelistof
minimal units, whereas temporal adverbs that potentially convey the same infor-
mation (“Yesterday,”) are not included. One reason for this decision might be that
those researchers are working on corpora based on journal texts, where temporal
expressions preferably are of the variety they include in the list; so the decision to
include one type of expression but not the other might be rather pragmatic than
theory-driven.
Other work has been less intention-oriented than that of Grosz and Sidner
and that in the rst paradigm, and consequently employed a different reasoning to
select units as elementary discourse units. One line of research that also dates back
to the mid-80s of the last century seeks to understand coherence in more general
terms than tied up with linguistics. In (Hobbs, 1985) and work that can be seen in
its tradition, like (Kehler, 2002), it is argued that coherence in text is by and large
a product of the capability of rational agents to understand the world as being
coherent. In this tradition, in its roots at least dating back to Hume (cf. (MacCormack
Calkins, 1913)), what is related by the rational mind are events or states of affairs.
Consequently, what counts as a minimal unit in these accounts are expressions
that can serve to convey states or events, or in short eventualities. A first class
citizen here is the clause again, and once again with suitable restrictions excluding,
e.g., restrictive relative clauses. For different reasons, Asher and Lascarides (2003)

Rhetorical structure 
and Reese et al. (2006, 2007) working in the Segmented Discourse Representation
Theory (sdrt) paradigm too consider the expression of eventualities to be the
decisive criterion for individuating minimal units.
As is well known from work in Montague grammar and elsewhere, it is all too
easy to coerce the type of expressions to that of an eventuality. In fact, the paper by
Jerry Hobbs in this collection (see 0.5) focuses on going below the clause level as
minimal units. Hobbs there points out that he does think that even single words
potentially express eventualities. It seems there is a thin line between raising types
to that of an eventuality too easily and missing out on sub-clausal constituents that
in fact are rhetorically interesting.
Yet another line of research different from the intention-oriented and the
eventuality-based ones can be seen in processing-based accounts. One of the
first candidates there, and also rooted in the mid-1980s is the work of Polanyi
(1986, 1988); but also the work by Webber (2004) can be seen in that tradition.
Polanyi’s ldm most closely mirrors the incremental nature of text processing in
that a discourse tree is built by adding sentences as elementary units to the existing
representation of the text so far perceived. Sentences obviously are larger units
than clauses since they potentially consist of multiple clauses (matrix, relative
clauses, complement clauses). Both the ldm and Webber’s d-ltag suggest exten-
sions to sentential syntax to model discourse structure, and thus there seems to be
no need for sub-sentential segmentation.
Both ldm and d-ltag — at least as concerns segmentation — thus seem to
follow the opposite strategy than that pursued by Hobbs in his contribution in the
present volume and reserve the rhetorical importance to larger units.
As a summary to the above approaches to segmentation, it seems that all
accounts agree on a core set of units (main clauses that form sentences) that should
be treated as elementary units, whereas there is large disagreement as to what else
should be considered a unit in discourse.
0.3â•… Rhetorical relations
In Section 0.2 various views on how to split up discourse were reported. The
present section is concerned with putting Humpty Dumpty together again: it
is agreed among linguists that coherent discourse should be represented as a
connected structure where each segment is connected to the rest by rhetorical
relations. Islands in the representation of the analysis of a text are dispreferred
and viewed as either a sign of incoherence of the discourse under analysis, faulty
analysis itself or some lack in descriptive power in the inventory of rhetorical
relations.

 Peter Kühnlein
In what follows in this section, the accounts used to introduce segmentation
strategies in the order chosen in Section 0.2 will be taken up in turn again and the
rhetorical relations employed by those accounts will be sketched.
On the intention-oriented side, Grosz and Sidner (1986) employ a surpris-
ingly small set of rhetorical relations. In their seminal paper they mention only
two of them, one being dominance (the dominated discourse unit serves to achieve
the goal of the super-ordinate) and the other satisfaction precedence (the preceding
goal has to be achieved before the next can be achieved). The authors are aware
of the fact that in, e.g., the work by Mann Thompson (ultimately published in
Mann and Thompson (1987a), but circulating in various grey versions before-
hand) a much larger number of rhetorical relations are discussed. However, since
for Grosz and Sidner primacy is on intentions and their relations to each other
rather than on textual realizations of intentions, they can claim that dominance
and satisfaction-precedence are sufficient for the description of rhetorical struc-
ture and the specific relations between textual units derivable from structure and
intention content.
Grosz and Sidner (1986) explicitly set up their account for construction
dialogues; given the goal orientation of that dialogue type, it seems that the inven-
tory consisting of dominance and satisfaction-precedence suffices to describe the
intentional structure of dialogues from that domain. This might be questioned,
however, in a more general domain, where a putative task structure (if there is any)
might be not as tightly bound to discourse structure. On the other hand, it has to
be said that Grosz and Sidner (1986) do not deny the existence of more relations
between discourse purposes. The claim, it seems, is just that the set suffices for the
analysis of the given type of discourse.
As mentioned, Mann and Thompson (1987a, 1988b) posit a much larger
set of discourse relations that should be used to describe the functional role
of elementary units as recognized by the analyst: according to that classifica-
tion, one unit might, e.g., elaborate on another, or units might form a list. There
are two main divides within the class of relations: the first divide concerns the
functional classification of relations: part of them are subject-matter relations
(reporting about facts), another part is presentational, employed to influence the
readers’ stance towards the (main or local) discourse topic. Both of these types
of relations can be realized in either of two ways (giving the second divide) —
connecting a less important part of discourse (a satellite) to a more important
one (called nucleus), or connecting nuclei to nuclei. The first type along the latter
divide is called mono-nuclear relation (or nucleus-satellite relation), the second
multi-nuclear.
One of the tests for nuclearity of discourse units is an elimination test: elimi-
nating nuclei from a text tends to render it incoherent, while eliminating satellites

Rhetorical structure 
tends to leave coherence intact. This property of nuclei has led Marcu (1996) to
posit the nuclearity principle for rst, claiming that spans of texts are connected by
a relation iff their nuclei are. (A consequence of that principle for the representa-
tion of discourse structure will be discussed in 0.4.) According to Matthiessen and
Thompson (1988), there is another indicator for nuclearity: they observed that
there is a high correlation between the status of being a nucleus in a text and of
being realized in a main clause just in case a rhetorical relation is present between
two syntactically related clauses. Syntactically subordinate clauses tend to realize
satellites in turn. As Matthiessen and Thompson (1988) warn, this is not a hard
and fast rule, but rather a tendency, and counter examples abound. There even
seem to be language specific discourse connectives that trigger an inversion in
nuclearity, like the dutch connective zodat (so that) which in a majority of cases
syntactically subordinates a nucleus to a satellite.
Bateman and Rondhuis (1994, 1997), and recently Stede (2008), systemati-
cally investigate rhetorical relations across different discourse theories and, for
rst’s nuclearity, propose not to tie the assignment of nuclearity to the presence
of certain rhetorical relations (i.e., to drop the divide between mono-nuclear and
multi-nuclear relations) and to view assignment of nuclearity as an effect of the
presence of other factors. This seems to be in line with the findings by Asher and
Vieu (2005) who claim something similar for an analogous divide among rela-
tions in sdrt. The insight that certain relations can be viewed as connecting nuclei
to satellites or satellites to other satellites seems also to be the rationale behind
the explosion of number of relations in the rst-flavor proposed by Carlson and
Marcu (2001), where multi-nuclear versions of relations formerly categorized
as mono-nuclear abound.
The discussion about the “right” relations for rst doesn’t seem to be settled
nor does it seem it has to be: Taboada and Mann (2006a) in their recent overview
over developments in rst propose that researchers in the paradigm should tailor
their own relations according to their specific needs for specific purposes.
The situation is different in sdrt, which, as mentioned, knows a similar divide
as the nucleus-satellite distinction in rst. sdrt knows thorough axiomatizations
of the discourse relations that are employed. These relations take the semantic
representations of the minimal units and join them in either of two ways: by
coordinating a unit to a preceding one, or by subordinating one unit to another.
The nature of the relation involved (subordinating or coordinating) has influence
on the possibilities where subsequent units can be attached: if the last relation
involved was coordinating, then the constituent to which the last unit was related
by it is blocked for attachment. If the last relation, on the other hand, was subordi-
nating, then both the last unit and the one to which it was attached are available.
These constraints on attachment points for new discourse units give rise to what is

 Peter Kühnlein
called the Right Frontier Constraint (rfc), first postulated by (Polanyi, 1986). One
effect of the rfc is to limit anaphoric accessibility: antecedents are said to be only
available for (pronominal) anaphoric uptake if they occur in a unit that is on the
right frontier.
Asher and Vieu (2005) re-examine the distinction between the two classes of
relations and suggest that the question whether, e.g., a cause-relation is subordinat-
ing or coordinating depends in part on the information structure exhibited by the
units that are related. Certain information structural configurations in the units
can lead to anaphoric accessibility of discourse referents whereas truth semanti-
cally equivalent variants of that information structure makes them inaccessible.
This fact can be accounted for if it is assumed that the information structure at
least in part can change the way a unit is attached to preceding discourse.
sdrt draws another distinction between discourse relations that resembles
the distinction between presentational and subject-matter relations in rst: many
of the discourse relations are content-level relations which are similar to subject-
matter relations, whereas other relations bear more resemblance to presentational
relations, like text-structuring, cognitive-level discourse relations or metatalk rela-
tions. Interestingly, the so-called satisfaction scheme for veridical relations holds
for some, but not all, relations of either variety. The latter scheme tells that two
(representations of) discourse units connected by a relation are part of the inter-
pretation of a discourse just in case the interpretations of the units are and the
interpretation of the relation is. Whereas a veridicality criterion like that is to be
expected for content-level relations, it is not so clear that a relation like parallel
(a text-structuring relation) should have that property. None of the cognitive-level
relations are veridical, though.
Just like sdrt and the account of Grosz and Sidner (1986), but unlike rst,
the account of discourse relations given by Hobbs (1985); Hobbs et al. (1993) and,
more or less based on it, Kehler (2002) is an attempt to give a principled way to
define the ways units of discourse are combined to form larger units. Hobbs et al.
(1993) distinguishes four classes of discourse relations: some that are inferred to
hold because the units that are connected are about events in the world (like casual
relations), others that relate what was said to an overall goal of the discourse, again
others that relate a unit to the recipient’s prior knowledge (e.g., backgroundâ•›) and
finally “expansion” relations (like contrastâ•›).
0.4â•… Structures and their properties
So, both what counts as minimal units and in which ways they can be combined by
rhetorical relations are matters of dispute in discourse theory. Given this situation,

Rhetorical structure 
it can be expected that there is also no consensus on which structures discourse
can be expected to have.
The expectation is confirmed by the literature. Whereas many researchers —
e.g., Polanyi (1986, 1988, 2001), Grosz and Sidner (1986, 1998), Mann and Thompson
(1987b, 1988a), Taboada and Mann (2006b) — assume that trees suffice to model
the rhetorical structure of dialogue, there is an increasing number of theorists that
doubt this assumption for a variety of reasons.
Prominent among the latter are Wolf and Gibson (2005, 2006) who argue for
a much less constrained type of graphs for the description of rhetorical structure.
The chain graphs they postulate as adequate for the description of rhetorical struc-
ture feature all kinds of violations of tree structure: they posit nodes with multiple
parents, crossing edges, and in general graphs without a distinguished root node.
Their strongest constraint on structures seems to be connectedness and acyclicity.
These graphs are capable of describing all kinds of relations between elementary
units; a closer look at their annotation manual and the set of relations they employ
reveals that this seeming strength is a real weakness too: the set of relations Wolf
and Gibson (2005, 2006) employ is a mixed bag, mostly taken from (Hobbs, 1985;
Hobbs et al. 1993), but considerably modified and enriched with some relations
from (Carlson Marcu, 2001). The annotation manual requires analysts to anno-
tate not only rhetorical relations used for combining minimal units, but also coref-
erence relations and other cohesive devices that can be present within minimal
units as well. Their first step of analysis, grouping, actually consists in connecting
units that are related by cohesive links. Only after that very step rhetorical relations
are applied to the units — alas not to the units connected by the first step. Thus, it is
no wonder that crossing dependencies and nodes with multiple parents abound in
the analyses presented in Wolf and Gibson (2005, 2006). Knott (2007) questions the
statistics Wolf and Gibson (2005, 2006) perform on their data, claiming that the
small percentage of tree violations that can be tied to the special relations intro-
duced by Wolf and Gibson in their evaluation of their data is implausible. I don’t
think so: rather, Knott’s critique seems to set in too late. The true reason for the
high amount of tree violations does not lie in the special relations, but in the
conflation of levels of analysis.
Another line of attack on tree structures as the adequate description for rhe-
torical structures can be found in (Danlos, 2004, 2008). Danlos compares the
generative capacity of a comparatively unrestricted formalism (an extension of
Mel’c̆uk’s (Roberge, 1979) dependency syntax to discourse) with those of rst and
sdrt. She derives all the structures that can be generated by either formalism and
tries to find discourses that exhibit the respective structure. Her benchmark for-
malism generates directed acyclic graphs (dags), whereas rst and sdrt generate
trees. According to Danlos analysis, rst undergenerates (is not complete), her

ïœ±ïœ° Peter Kühnlein
benchmark formalism overgenerates (is not correct), and sdrt is closest to being
bothcompleteandcorrect, with theexception being a few structures that can not be
described as sdrt-trees. Danlos argument to my mind has two drawbacks, though
it is admirably ingenious. First, it should be evaluated on corpus data instead of
relying on constructed discourse for confirmation. This is mainly a precaution
against an overreliance on intuitions, of course. The second point is a bit stronger: the
reconstruction of rst mainly — based on (Marcu, 1996) and (Carlson Marcu,
2001) — seems to contain too strong an interpretation of the nuclearity principle
that leads to the assumption of graphs that are not in accordance with most other
work in rst. So, rather than raising an argument against general rst assumptions,
her attack is directed against a very idiosyncratic version of rst.
But both the account of Wolf and Gibson (2005, 2006) and of Danlos (2004,
2008) are under active discussion, and until there is conclusive evidence to the
contrary, it has to be assumed that there are strong arguments against general tree-
hood of discourse structure. A weaker warning comes from Webber (2001) and
Lee et al. (2008): these authors caution that although most discourse can be mod-
elled as trees, there might be certain cases where a departure from tree structures
is required. So the warning would be to give up the general claim in favour of a
rule of thumb with dened exceptions.
It seems that the question of how rich a structure has to be assumed for the
description of discourse these days is more hottly debated than ever. The papers
in the present volume will help to solve or focus in this debate by contributing
insights in the fundamental questions that have to be answered.
0.5â•… About the papers
Jerry R. Hobbs: Clause-Internal Coherence
As was discussed in Section 0.2, there is no unanimity about the size or gen-
eral characterization of elementary discourse units, just as there is no agreement
on the definitions of relations between them. In his contribution to this volume,
Hobbs extends his account, e.g., from (Hobbs, 1985), to cover coherence at a
sub-clausal level.
Henk Zeevat: Optimal Interpretation for Rhetorical Relations
Zeevat argues that rhetorical relations can be reconstructed from general
optimality theoretic (ot) assumptions. He gives a comprehensive introduction to ot,
with special emphasis on the constraints *new, relevance, faith and plausi-
ble. He then continues to demonstrate how a range of rhetorical relations can be
derived from a certain ordering of these constraints; most importantly, *new and

Rhetorical structure ïœ±ïœ±
relevance tend to introduce rhetorical structure defaults, with plausible being a
filter over the generated relations. This account of coherence relations is in marked
contrast to accounts such as that of Hobbs (1985) or Asher and Lascarides (2003).
Ekatarina Jasinskaja: Modelling Discourse Relations by Topics and Implicatures:
The Elaboration default
Jasinskaja argues in her paper for the position that discourse relations can be
inferred by utilising underlying pragmatic principles such as topic continuation
and exhaustive interpretation as defaults. In the absence of linguistic markers that
make one more inclined to infer a different relation, she opts for Elaboration
as one of the default relations, since it best obeyes both principles, i.e., does not
induce topic shifts and at the same time add information to the topic at hand.
Maja Bärenfänger, Harald Lüngen, Mirco Hilbert Henning Lobin: The role
of logical and generic document structure in discourse analysis
The authors of this contribution propose to add two descriptive levels to the
local rhetorical analysis of discourse structure: the logical structure (like title,
paragraph etc.) and the genre specific structure (introduction, method). Structure
at these levels is usually not explicitly signalled, yet conventionalized, and can thus
be used to guide (automatic) parsing of texts. The authors strive to clarify which
cues and constraints can be observed at these levels of discourse and demonstrate
their utility for automatic text processing.
Pascal Amsili Claire Beyssade: Obligatory Presupposition in Discourse
Presupposition triggers have been considered obligatory under certain condi-
tions by a variety of authors. One of the conditions that was deemed necessary in
previous work was that the triggers are additive particles, like too. Amsili Beyssade
argue that this condition is not a necessary one, but that obligatoriness is the case
for triggers that have no asserted content. (Too being but one of them.) They give
a general explanation for the apparent sensitivity of this class of triggers to discourse
relations and provide a formalization in terms of an sdrt update mechanism,
building on Asher and Lascarides (1998).
Ann Copestake Marina Terkourafi: Conventionalized speech act formulae —
from corpus findings to formalization
Copestake Terkourafi present an account to the semantics and pragmatics
of conventionalized speech acts which renders the contribution of the illocution-
ary force as an addition to the compositional semantics of the utterance. They
motivate their account with examples from a corpus of Cypriot Greek and formalize
it within the framework of hpsg. They show how their account leaves the possibility
for a literal interpretation of conventionalized speech act formulae open, thus
opening the possibility to react to them in a variety of ways.

ïœ± Peter Kühnlein
Philippe De Brabanter: Constraints on metalinguistic anaphora
De Brabanter reports the results of his research into a specific kind of referring
expressions, where the referent itself is a linguistic object (like in “â•›‘Boston’ is disyl-
labic”). He draws a number of distinctions among those expressions, arguing that
the class of metalinguistic anaphora referring back to expressions that themselves
do have non-linguistic referents are especially interesting for a number of reasons.
Cyril Auran Rudy Loock: Appositive Relative Clauses and their Prosodic
Realization in Spoken Discourse: a Corpus Study of Phonetic Aspects in British
English
In their contribution to the volume, Auran Loock argue that differences in
pragmatic functions fulfilled by appositive relative clauses are correlated to differ-
ences both in morphosyntactic and semantic characteristics and prosodic features.
The latter mainly concern intonation, rhythm and intensity. The data they use are
extracted from a corpus of spoken British English.
Bibliography
Asher, N. Lascarides, A. (1998). The semantics and pragmatics of presupposition. Journal of
Semantics, 15: 239–99.
Asher, N. Lascarides, A. (2003). Logics of Conversation. Cambridge University Press.
Asher, N. Vieu, L. (2005). Subordinating and coordinating discourse relations. Lingua, 115:
591–610.
Bateman, J. Rondhuis, K.J. (1994). Coherence relations: analysis and specification. Technical
Report R1.1.2: a,b, DANDELION Esprit Basic Research Project 6665.
Bateman, J. Rondhuis, K.J. (1997). Coherence relations: towards a general specification.
Discourse Processes, 24: 3–49.
Benz, A. Kühnlein, P., editors (2008). Constraints in Discourse, volume 172 of Pragmatics and
Beyond new series. John Benjamins.
Butterworth, B. (1975). Hesitations and semantic planning in speech. Journal of Psycholinguistic
Research, 4: 75–87.
Carlson, L. Marcu, D. (2001). Discourse tagging manual. Technical Report ISI-TR-545, ISI.
http://www.isi.edu/marcu/discourse/tagging-ref-manual.pdf.
Chafe, W.L. (1980). The Pear Stories, volume 3 of Advances in Discourse Processes, chapter The
Deployment of Consciousness in the Production of a Narrative, pages 9–50. Ablex.
Danlos, L. (2004). Discourse dependency structures as constrained dags. In Strube, M. Sidner, C.,
editors, Proceedings of the 5th SIGdial Workshop on Discourse and Dialogue, pages
127–135.
Danlos, L. (2008). Strong generative capacity of rst, sdrt and discourse dependency dags. In
Benz, A. Kühnlein, P., editors, Constraints in Discourse, pages 69–96. John Benjamins.
Grosz, B. Hirschberg, J. (1992). Some intonational characteristics of discourse structure.
In Proceedings of the International Conference on Spoken Language Processing, pages
429–32.

Rhetorical structure ïœ±
Grosz, B.J. Sidner, C. (1986). Attention, intentions, and the structure of discourse. Computa-
tional Linguistics, 12(3): 175–204.
Grosz,B.J.Sidner,C.(1998).LostIntuitionsandForgottenIntentions.InWalker,M.A.,Joshi,A.K.,
Prince, E.F., editors, Centering Theory in Discourse, pages 39–51. Clarendon Press.
Hirschberg, J. Nakatani, C.H. (1996). A prosodic analysis of discourse segments in direction-giving
monologues.InProceedingsoftheAnnualMeetingoftheAssociationforComputationalLinguistics,
pages 286–93, Santa Cruz.
Hirschberg, J. Pierrehumbert, J. (1986). The Intonational Structuring of Discourse. In
Proceedings of the 24th Annual Meeting of the Association for Computational Linguistics,
pages 136–44.
Hirschberg, J. Pierrehumbert, J. (1992). The intonational structuring of discourse. In
Proceedings of the 24th Annual Meeting of the Association for Computational Linguistics,
pages 136–44. ACL.
Hobbs, J. (1985). On the coherence and structure of discourse. Technical Report 85-37, CSLI.
Hobbs, J., Stickel, M., Appelt, D., Martin, P. (1993). Interpretation as abduction. Technical
report, SRI International.
Kehler, A. (2002). Coherence, Reference and the Theory of Grammar. CSLI.
Kempen, G., editor (1987). Natural Language Generation. Number 135 in NATO Advanced
Science Institutes—Applied Sciences. Martinus Nijhoff Publishers.
Knott, A. (2007). Review of “Coherence in Natural Language: Data Structures and Applications”
by Florian Wolf Edward Gibson. Computational Linguistics, 33(4): 591–5.
Lee, A., Prasad, R., Joshi, A., Webber, B. (2008). Departures from tree structures in discourse:
Shared arguments in the penn discourse treebank. In Benz, A., Kühnlein, P., Stede, M.,
editors, Proceedings of CID III.
Lüngen, H., Puskás, C., Bärenfänger, M., Hilbert, M., Lobin, H. (2006). Discourse Segmentation
of German Written Texts. In Advances in Natural Language Processing, volume 4139 of Lecture
Notes in Computer Science, pages 245–56. Springer.
MacCormack, T.J. Calkins, M.W., editors (1913). Hume, David: An enquiry concerning human
understanding and selections from A treatise of human nature, volume 7 of Bibliotheca
philosophorum. Meiner.
Mann, W.C. Thompson, S.A. (1987a). Rhetorical Structure Theory: A Theory of Text Organization.
Technical Report RS-87-190, Information Sciences Institute, Los Angeles, CA.
Mann, W.C. Thompson, S.A. (1987b). Rhetorical Structure Theory: Description and
Construction of Text Structures. in: (Kempen, 1987). pp. 85–95.
Mann, W.C. Thompson, S.A. (1988a). Dialogue Games: Towards a functional theory of text
organization. Text, 8(3): 243–81.
Mann, W.C. Thompson, S.A. (1988b). Rhetorical Structure Theory: Toward a functional
theory of text organization. Text, 8(3): 243–81.
Marcu, D. (1996). Building Up Rhetorical Structure Trees. In The Proceedings of the Thirteenth
National Conference on Artificial Intelligence, pages 1069–74, Portland, Oregon.
Matthiessen, C. Thompson, S.A. (1988). The structure of discourse and ‘subordination’. In
Haimann, J. Thompson, S.A., editors, Clause Combining in Grammar and Discourse,
volume 18 of Typological Studies in Language. John Benjamins Publishing Company.
Polanyi, L. (1986). The linguistic discourse model: Towards a formal theory of discourse structure.
Techn. Report TR-6409, BBN Laboratories Inccap.
Polanyi, L. (1988). A formal model of the structure of discourse. Journal of Pragmatics,
12: 601–38.

Random documents with unrelated
content Scribd suggests to you:

Long Wood (Lochanhead), 174
Dumfriesshire—
Caerlaverock, 180
Auchencreath, 175
Dalswinton, 183
Closeburn, 182
Drumlanrig, 183
Sanquhar, 184
Kirkconnel, 184
Polveoch, 184
Kello Water, 184
Glen Aylmer, 184
Glen Wharry, 184
Bale Hill, 186
Annandale, 184
Lochmaben, 175
Burnswark, 184
Corrie, 185
Fin M‘Coul, 43
“Fire Spangs of Faustus,” 16
Funeral festivities (“Gallovidian Encyclopædia”), 232
Funeral refreshment (Draigie), 234
Funeral rites and customs, 236
Funeral “services,” 225
G
Galdenoch Tower, 245

“Galloway Gossip,” 166
Galloway Mansion near Castle-Douglas, Ghostly story of, 273
“Galloway Register,” 26
“Galloway Traditions,” 26
Galloway, Western, Traditions of, 22
Gap’s Mill, Glencairn, 283
Garryhorn, 36
Gatehouse, 262
General Assembly (Condemnatory Acts), 68
“Gentle Shepherd” (extract from), 59
Ghost-lore and Haunted Houses, 244
Ghost Legends of the South-west of Scotland (arranged in their
order, from West to East)—
Dunskey Castle, 244
Galdenoch Tower, 245
“Lodnagappal Plantin’,” 248
High Ardwell, 248
Auchabrick House, 250
Cardrain House, 251
Tirally, 251
Glenluce, 252
Provost Coltran (Drummorall), 252
Packman’s Grave (Bladnoch), 253
Claunch, Sorbie, 254

Whithorn, 254
Craigdhu, Glasserton, 255
Church of Kirkmaiden, 256
Auchenmalg Barracks, 257
Craighlaw House, 257
Machermore Castle, 258
Creetown, 262
Kirkdale Bridge, 263
Glenlee, Dalry, 263
Bogle-Hole, Dalry, 267
Moor of Corsock, 267
Buckland Glen, 269
Ringcroft of Stocking, 272
Mansion House near Castle-Douglas, 273
Wood Forester’s, Dalbeattie, 274
Laird o’ Coul’s Ghost, 300, 344
Kirkbean—
Murder Fall, 274
Derry’s How, 274
Farm-house, 274
Howlet’s Close, 275
Three Cross Roads, 275
Near Newabbey, 276
Minorite Friary, Dumfries (1306) and Caerlaverock Castle, (1358),
276
Solway legend of the passing of “Aul’ Lag,” 278
Coach legend of passing of William Duke of Queensberry
(Drumlanrig), 281
Drumlanrig Castle, 282
Tynron Doon, 282
Glencairn—
Auchenstroan, 283
Marwhirn, 283
Pentoot, 283
Gaps Mill, 283
Nut Wood, 283

Jarbruck Bridge, 283
Kirkland Bridge, 283
Sanquhar Castle, 283
Littlemark, Sanquhar, 284
Abraham Crichton’s Ghost, 285
Poldean, Wamphray, 287
Spedlins Tower, 288
Jardine Hall, 290
Knockhill, 293
Orchard, Hoddom, 294
Bonshaw Tower, 294
Blackett Tower, 294
Kirkconnel Hall, 295
Todshawhill, 296
Lowland Manor House, 298
Gilchristland, 50
Gilroanie, 270
“Girzie M‘Clegg,” 17
Glasserton, 165, 215
Glencairn, 283
“Glencairn Kate,” 17
Glencaple Quay, 199
Glenkens, 19
Glenkens, twenty years’ holidaying in (Blacklock), 265
Glenlee House (ghost narrative), 263

Glenluce, 13, 14
Greenmill (Caerlaverock), 209
Grennan, The, 25
Grierson, John, of Lag (funeral expenses of), 227
Grierson of Lag, Sir Robert (funeral expenses of), 229
Grierson of Lag, Sir Robert (funeral legend), 230
Grierson of Lag (Solway legend of his “passing”), 278
Grose’s “Antiquities of Scotland,” 289
“Gyre Carline,” 8
H
Hallowmass, 2
Hallowmass Rade, 3
Hannayston, Witch of, 17
Harper’s “Rambles in Galloway,” 17
Hay of Park, 60
Heron, Robert (Journey through Western Scotland), 54
High Ardwall (white woman apparition), 248

Holm Glen (Dalry), 275
Howlet’s Close (Kirkbean), 275
“Hydrostatics,” Sinclair’s, 300
I
“Il Penseroso” (extract from), 186
Inshanks Moor, 29
Irvings of Hoddom, 293
J
James VI. of Scotland, 67
Jarbruck, 283
Jardine’s of Applegarth, 289
Jardine Hall, 290
“Jean o’ the Howff” (Rerwick), 45
“Jock o’ the Horn,” 182
K
Kain Bairns, 7

“Keekafar” (Kirkmaiden), 155
Kells, 35
Kells Rhynns, 36
Keltonhill, 40
Kenmure (Stoneykirk), 157
Kenmure Castle (Dalry), 269
Killymingan (Kirkgunzeon), 105
Killumpha Farm (Kirkmaiden), 204
Kilmeny (Jas. Hogg), 146
Kincaid, John (Witch-pricker), 70
King’s Croft of Stocking, 63
Kirkdale Bridge, Ghost of, 263
Kirkdale House, 262
Kirkmaiden, 22, 29, 151
Kirkmaiden Church, 30
Kirkmaiden, Legend of, 256
Kirkmaiden Witches, 29, 32, 98
Kirk-session (Borgue) examination for alleged fairycraft, 159

Kirkpatricks of Closeburn, 214, 227, 231, 284
Kirkpatrick, Roger, 277
Kirkpatrick Sharpe, Charles, 288
Kirkwaugh (Bladnoch), pedlar’s ghost at, 253
Kippford, 274
Kirkennan Woods (Dalbeattie), 199
Kirkland Bridge (Glencairn), 283
Knockhill Mansion (tragedy at), 293
Knocknishy (Whithorn), 185
Knocksheen (Dalry), 17
L
Lady Ashburton, 267
Laird o’ Coul’s Ghost, 344
Langhill Fairy, The, 166
Lapps or Finns, 149
Latewake, 223
Law’s Memorials, 287

“Lay of the last Minstrel” (extract from), 16
Liethin Hall, 187
Leswalt, 245
Levitical Law, 68
Library of Michael Scott (list of works), 16
Lichts before death, 209
Lindsay, James (Caerlaverock tragedy), 277
Little Cocklick (Urr), 101
Littlemark Farm, Sanquhar, Ghostly appearance at, 284
Locharbriggs Hill, 3
Lochar Moss, 8
Loch Doon, 36
“Lodnagappal Plantin,” Apparitions at, 248
Logan, 24, 25
Logan Mill, 31
Lord Crichton (6th), 284
Lord Glenlee, 263
Lords of Sanquhar, 284

Lord Stormonth, 227
Lotus Hill (Kirkgunzeon), 173
Loup o’ the Grennan, 151
Low Curghie (Kirkmaiden), 24
Luce, 13, 15
Luce Bay, 215
Lykewake, 223
M
Machars of Galloway, 33
Machermore Castle, Legend of, 258
Maggie’s gate to Gallowa’, 13
Mainsriddel, 274
“Maggie o’ the Moss,” 6, 17, 21
“Mak’ Siccar” (tragedy, Dumfries), 278
Manor House in Lowlands (story of apparition), 298
Manxman’s Lake, 270
March Moon, 55

Marshall, Rev. Mr (Kirkmaiden), 97, 248
Marwhirn, 283
Millar, Mary (alleged witch), 74
Mary Queen of Scotland (Act against witchcraft), 66
Master of Logan (Allan Cunningham), 19
Maxwell of Carriel (Carzield), 227
Maxwell of Dalswinton, 188
Maxwells of Monreith (successors to M‘Cullochs), 214
Maxwell, Thomas (Laird of Coul), 301
Maxwell, Jean, trial of (for pretended witchcraft), 98
Maxwell, Jean (copy of title page of publication of trial), 110
Meg Elson (Kirkmaiden witch), 32
Meg Elson’s Elegy, 32
Meg Macmuldroch (Galloway witch), 62
Melrose Abbey, 16
Michael Scott of Balwearie, 15
Mochrum Parish (extravagant funeral expenditure), 226
Moffat Churchyard, 213

Monkland Shore, 44
Monreith House, 161
Moor of Corsock (ghost of headless piper), 267
Moor of the Genoch, 248
Moor Kirk of Luce, 13
Mort-cloth (use of), 239
Mountsallie (Rhinns), Witchcraft at, 57
Muirhead, Dr James, 107
Mull of Galloway, 149
Murder Fall (Kirkbean), 274
Myrton Mound (fairy legend), 161
M‘Cullochs of Myrton, 214
M‘Culloch, Sir Godfrey, 151
M‘Millan Cup, 195
M‘Milligan of Dalgarnock, 283
N
“Necromancy,” 16

Newabbey, Witchcraft at, 10
Newabbey (ghost of lady in white), 276
Nicholas Grier (witch of Hannayston), 17
Nick o’ the Balloch, 13
“Nithsdale Minstrel” (poetical collection), 34
Nith, 51, 189
Nut Wood, Maxwelton (Moniaive), 283
Nicholson, Wm., poet (fairycraft examination, recollection by his
mother), 159
O
“Old Church life in Scotland” (Edgar), 237
Old Hall at Ecclefechan, Ghost at, 295
Old House of Park, 61
Old John Orr (Carlyle reminiscence), 293
Old Meg of Twynholm (reputed witch), 43
Old Red Cap (ghost of Blackett Tower), 294
Old Turnpike House, Dumfries, 231
Orchard, Hoddom (laying of ghost), 294

Osborne, “Maggie” (Wigtownshire witch), 11
P
Packman’s Grave (Bladnoch), 258
Palmallet (Whithorn), 96
Palnackie, 199
“Passing Bell” (custom of ringing), 241
Passing Bell (reference in “Book of Galloway”), 243
Patiesthorn, Legend of, 269
“Pawky Auld Kimmer,” 65
Pentoot (Glencairn), 283
“Philosophy of the Devil,” 16
Picts, 148, 149
Poldean, Wamphray (ghost reference), 287
Portankill (fairy haunt), 149
Porteous, ghost of, at Spedlins Tower, 289
Portencockerie Bay (fairy haunt), 156
Port Logan, 31, 156

Portpatrick, Legend of, 245
Port-William, 254
Presbytery of Penpont (warning regarding burial festivity abuse), 234
Prestonmill, 274
“Pricking” of Witches, 70
“Prince of Darkness” (and witch revelry), 8
Privy Council Commissions (to try cases of witchcraft), 71
R
Rab’s Howff (Rerwick), 45
Ray’s Itinerary (Dumfries), 242
Red Comyn, 277
Rerwick, 44
Rerwick Apparition, 272, 321
Rhinns, 25
Rhonehouse, 40
“Riddling in the Reek,” 166
“Ridden post by a witch,” 5

Ringdoo Point, 15
Ringcroft of Stocking, 272
Ringcroft of Stocking, site of, 300
Robert the Bruce, 36
“Robin Goodfellow,” 186
Roodmas, 176
Rotten Row (Whithorn), 33
S
Sanquhar, 50
Sanquhar Castle (ghostly legends), 283
Sanquhar, History of (Simpson), 184, 285
Sanquhar Kirkyard, 240
“Satan’s Almanac,” 16
“Satan’s Invisible World,” 300
Scots Money, 227
Shaws of Craigenbay and Craigend, 35
Shawn (Stoneykirk), 185

Shennaton (Bladnoch), 64
Shinnel Water, 283
Shirmers, 269
Sin-eating, 218
Sir Chesney Shaw, 35
Sir Walter Scott, 16, 244
Slip Coffins, 237
Solway Firth, 8
“Soothsayers’ Creed,” 16
Spell-casting, 60
Spedlins Tower, Ghost of, 288
Spedlins Tower Bible, 291
St. Ninian, 39
Stake Moss, Sanquhar, 239
State and Church (action against witchcraft), 22
Stewartry of Kirkcudbright, 151
Stoneykirk, 36, 248
Suicides, Burial of, 239

Surprising Story of the Devil of Glenluce, 299, 302
Sweetheart Abbey, 2, 10
T
Tam o’ Shanter, 6, 17
Telfair, Alexander (Minister of Rerwick), 272
Three Cross Roads (Kirkbean), 275
Tirally (Kirkmaiden), 56
Tirally, Ghost at, 251
Todshawhill, Bogle of, 296
Tolbooth of Kirkcudbright, 108
Tongland, 16
Tower of Craigend, 35
Traditional Witchcraft described, 1
Train, Joseph (account of funeral superstitions), 236
True account of an apparition in Ringcroft, parish of Rerwick, 299,
321
Tynron, 49

Tynron Doon, Spectre of, 282
U
“Unique Traditions of the West and South of Scotland” (Barbour), 35
Upper Nithsdale, 50
W
“Warlock Feckets,” 55
“Walpurgis” (witch festivals), 8
Warnings, accounts of from—
Caerlaverock, 209
Closeburn, 214
Corrie, 2
Craigdarroch, 214
Dumfries, 213
Glencairn, 210
Kirkmaiden (in Fernes), 214
Moniaive, 208
Tynron, 209
Waterside Hill (Dalry), 19
Water of Urr, 207
“Waulking” the dead, 219
Walter de Curry, 244

Well of the Co’ (Kirkmaiden), 150
White Loch of Myrton, 161
Whithorn, Old Manse, 254
Whinnieliggate, 40
Whithorn (similar legend to Tam o’ Shanter), 33
White Lady of Machermore, 258
“Witch Cake,” 9
“Witch Chronicle, The,” 16
Witches Gathering, 3
Witch Marks, 8, 70
Witch Narrative, 21
Witch Narrative (Southern Kirkcudbrightshire), 40
Witches Sabbath, 7
Witches’ Stairs (Crawick), 50
Witches’ Rocks (Portpatrick), 36
William, Duke of Queensberry (legend of ghostly coach), 281
Witchcraft, proceedings against, in Galloway—
Kirkcudbright (Presbytery, 1662), 72
Kirkcudbright, 1671, 72
Dalry (Kirk-session, 1696), 72

Dalry (Kirk-session, 1697), 73
Kirkcudbright, 1701, 82, 86, 87
Twynholm, 1703, 87
Urr (parish of) 1656, 91
Kirkpatrick-Durham (parish of), 92
Carsphairn (parish of), 93
Minnigaff (parish of), 93
New Luce (parish of), 96
Whithorn (parish of), 96
Kirkmaiden (parish of), 97
Maxwell, Jean, trial of (pretended witchcraft), 98
Dumfriesshire (proceedings in)—
Burgh of Dumfries, 1657, 111
Kirk-Session of Dumfries, 1658, 111
Dumfries (official information regarding the judicial burning of
nine women), 112
Dumfries (attendance of clergy at the burning), 115
Dumfries (resolution against Janet Burnes, alleged witch), 115
Dumfries (warrant of execution against two alleged witches),
116
Dumfries (last trial for witchcraft in Scotland, Elspeth Rule), 117
Dumfries (Presbytery of—Southern district), 118
Caerlaverock, Kirk-session records, 118
Irongray, Kirk-session records, 120
Irongray Parish (traditional account of witch punishment), 122
Closeburn Parish, 124
Penpont Presbytery, 131
Glencairn Kirk-session records, 132
Glencairn, Case of Alexander Deuart, 133
Durisdeer, 138
Torthorwald, 140

Wood Foresters’, Dalbeattie (scene of murder and ghost
appearance), 273
Warnings, 207, 208, 209, 210, 211, 212
Wraiths—
Seen at Balgreggan House, 205
Buittle, 199
Dalbeattie, 205
Glencairn, 201
Kirkmaiden, 204
Moniaive, 202
Wraiths (account of from “Gallovidian Encyclopædia”), 202
Wylliehole, Witch of, 53
Y
Yule, 278
Yule Candles, 219
Footnotes:
[1] The Well of the Co’, Kirkmaiden, once much celebrated for the
healing and medicinal properties of its waters.
[2] These berries make excellent preserves.

[3] Heather after being burned.
[4] “Confessions of Isobell Goudie.”
[5] Dwining.
[6] Shall be.
[7] Stubble.
[8] Kiln.
[9] Sighing.
[10] A famous haunt of witches in the parish of Rerwick.
[11] Extract from King James’s Daemonologie concerning Sorcery
and Witchcraft (1597):—
“The persons that give themselves to witchcraft are of two sorts, rich
and of better accompt, poore and of baser degree. These two
degrees answere to the passions in them, which the divell uses as
means to entice them to his service: for such of them as are in great
miserie and povertie, he allures to follow him, by promising unto
them great riches and worldly commoditie. Such as though rich, yet
burne in a desperate desire of revenge, he allures them by promises
to get their turne satisfied to their heart’s contentment.”
[12] “The witch mark is sometimes like a blewspot, or a little tate, or
reid spots, like flea-biting; sometimes also the flesh is sunk in, and
hallow, and this is put in secret places, as among the hair of the
head, or eyebrows, within the lips, under the armpits, et sic de
ceteris.” Mr Robert, minister at Aberfoill, in his Secret
Commonwealth, describes the witch’s mark—“A spot that I have
seen as a small mole, horny, and brown-coloured; through which
mark, when a large brass pin was thrust (both in buttock, nose, and
rooff of the mouth) till it bowed and became crooked, the witches,
both men and women nather felt a pain nor did bleed, nor knew the

precise time when this was being done to them (their eyes only
being covered).”—Law’s “Memorials,” ed. by C. Kirkpatrick Sharpe.
[13] The extreme penalty took two forms. The condemned were
either in the first place strangled or, to use an old expression,
“wirreit” and then burned; or, worse still, they were straightway
burned quick (alive).
[14] Thessr = Treasurer.
[15] Printed in Dumfries by his brother, Robert Rae, 1718.
[16] The Parish of Glencairn, Rev. John Monteith.
[17] Coshogle mansion-house or keep, belonging to the Douglases,
was situated on the hill overhanging the Enterkine burn, above the
farm-house of the same name. A marriage stone, built into a cottage
wall, is all that remains of the structure.
[18] Sir James Douglas of Parkhead, styled Lord Torthorwald as
having married the heiress of that barony, was afterwards run
through the body on the High Street of Edinburgh by a nephew of
Captain James Stewart, and died without uttering one word. On
clearing away the rubbish, which till lately covered the pavement of
the Chapel at Holyrood House, his tombstone was found, with this
mutilated inscription:—“Heir lyes ane nobil and potent Lord James
Douglas—and Cairlell and Torthorall wha mariet Daime Elizabeth
Cairlell, air and heretrix yr. of, wha was slaine in Edinburgh ye 14 day
of July, in ye yeir God 1608.”—Law’s Memories.
[19] Another theory associates the fairies with the dwarfish Lapps or
Finns who, driven out of their own country, settled in the outlying
districts of Scotland.
[20] The mother of William Nicholson the poet, a native of Borgue,
where her family had long been settled, and a woman of great
intelligence, often told that in her day there lived a man belonging to
Borgue parish whose mother and grandmother had been examined

before the Kirk-Session regarding his having been carried away by
the fairies.
[21] “Brownie” here synonymus with “Fairy.”
[22] Langhill (now Longhill), adjacent to the Rispain Roman Camp,
about a mile from Whithorn on the Glasserton Road.
[23] Roodmass: The festival of the finding of the Holy Cross (May
3rd).
[24] “When the mother’s vigilance hinders the fairies from carrying
her child away, or changing it, the touch of fairy hands and their
unearthly breath make it wither away in every limb and lineament
like a blighted ear of corn, saving the countenance, which
unchangeably retains the sacred stamp of divinity. The way to cure a
breath-blasted child is worthy of notice. The child is undressed and
laid out in unbleached linen new from the loom. Water is brought
from a blessed well, in the utmost silence, before sunrise, in a
pitcher never before wet; in which the child is washed, and its
clothes dipped by the fingers of a maiden. Its limbs, on the third
morning’s experiment, plump up, and all its former vigour returns.”—
Allan Cunningham, in “Cromek’s Remains of Nithsdale and Galloway
Song.”
[25] The leaden figure of a man connected with a cascade, once a
prominent feature of the gardens.
[26] Simpson’s History of Sanquhar.
[27] The “Brownie” of Scotland corresponds with the “Robin
Goodfellow” of England.

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Constraints In Discourse 2 Peter Khnlein Ed Anton Benz Ed

More Related Content

Similar to Constraints In Discourse 2 Peter Khnlein Ed Anton Benz Ed (20)

Recently uploaded (20)

Constraints In Discourse 2 Peter Khnlein Ed Anton Benz Ed