type,startTime,endTime,query,links_found,url,content_length,content,link_idx
tavily_search,0.0,3.533738613128662,"Which figure in the paper ""Identifying semantic role clusters and alignment types via microrole coexpression tendencies"" shows the hierarchical clustering of similarities in microrole coexpression?",5.0,,,,
page_scrape,3.7183144092559814,7.0980730056762695,,,https://www.ingentaconnect.com/content/jbp/sl/2014/00000038/00000003/art00002,1099.0,"# Identifying semantic role clusters and alignment types via microrole coexpression tendencies
In this paper, we illustrate a method for identifying clusters of semantic roles by cross-linguistic comparison. On the basis of data from 25 languages drawn from the ValPaL (Valency Patterns Leipzig) database, we show how one can visualize coexpression tendencies using quantitative methods (in particular, multidimensional scaling). Traditionally, the coexpression of semantic microroles (such as the breaker and the broken thing of the ‘break’ verb, the helper and the helpee of the ‘help’ verb, etc.) has been studied for particular languages, with generalized macroroles such as “agent”, “actor”, and “undergoer” being compared across languages in a next step. We set up a conceptual space of 87 microroles based on their coexpression tendencies, i.e. the extent to which they are expressed identically (via flagging and indexing) across our languages. The individual coding means (cases, adpositions, index-sets) can then be mapped onto this conceptual space, revealing broader alignment patterns.
",2.0
page_scrape,3.746093988418579,5.168972492218018,,,https://www.researchgate.net/publication/304460742_Identifying_semantic_role_clusters_and_alignment_types_via_microrole_coexpression_tendencies,12613.0,"# Identifying semantic role clusters and alignment types via microrole coexpression tendencies
. Seven microroles with associated coding sets in three different languages 7
Identifying semantic role clusters
and alignment types via microrole
coexpression tendencies
In this paper, we illustrate a method for identifying clusters of semantic roles by
coexpression tendencies using quantitative methods (in particular, multidimen-
sional scaling). Traditionally, the coexpression of semantic microroles (such as
languages in a next step. We set up a conceptual space of 87 microroles based on
their coexpression tendencies, i.e. the extent to which they are expressed identi-
ters can then be identied by studying cross-linguistic coexpression tendencies, i.e.
the ways in which the individual microroles cluster with respect to their coding
therst paragraph, our method assumes that the coexpression of roles is not ran-
and the servee are coexpressed in German (by Dative case). Accidental homony-
my may exist, of course, but repeated coexpression of the same notional elements
across many dierent languages must indicate similarity of meaning (e.g. Haiman
semantic-map method, applied to microroles of individual verbs and their coding
Identifying semantic role clusters and alignment types via microrole coexpression tendencies465
tic content. For example, the ‘agent’ might subsume the microroles ‘hitter’ and
(Van Valin 2005) (or hyperroles, Kibrik 1997).e way in which microroles can
the verb-specic level of microroles is that it has no language-internal generality at
for the similar behavior of ‘break’, ‘hit’, and so forth), but it also allows for cross-
In this paper, we only use microroles to compare languages.4While we are
scription and for comparison.e microroles that we use here are thus intended
3.From macrorole alignment to microrole coexpression
ment are the accusative type, where intransitive S is coded like A (= coexpressed
like P (= coexpressed with P) but dierently from A.is is generally represented
Identifying semantic role clusters and alignment types via microrole coexpression tendencies467
variation within single-argument verbs (agentive-patientive and similar systems,
coding, we need to extend the alignment or coexpression approach to a larger set
e coexpression of theve roles is rather dierent in English and German, as
be seen in Table 1, which compares the coding sets6of seven microroles in three
other languages. We see that only the ‘hitter’ and the ‘helper’ are coexpressed in all
Figure 2.Examples of macrorole (a.) and microrole (b.) argument alignments
are only coexpressed in one language (Icelandic). Comparing such pairs of micro-
roles and their coexpression or non-coexpression will be the basis for our study.
4.Microrole coexpression in 25 languages
Table1.Seven microroles with associated coding sets in three dierent languages7
microrole IcelandicHoocąk Chintang
Identifying semantic role clusters and alignment types via microrole coexpression tendencies469
Our 87 verb meanings can be found in Appendix 1 in a manner similar to how
than 5 languages) we only use 181 microroles for the current paper. Because the
verb meanings correspond closely across languages, the microroles of the verbs
which of the microroles are coded in the same way (i.e. coexpressed) within each
single language. Basically, the number of coexpressions between two microroles,
averaged over the 25 languages, provides an estimate of the similarity between the
metric, representing a semantic map of microroles across languages (cf. Cysouw
In this paper, the metric on microroles was dened as follows. First, when the
coding set of a microrole consists of two coding elements (aag and an index), we
microrole is coded by Nominative case (aag) and Subject agreement on the verb
(an index-set).e similarity between two microroles within a language was then
of microroles, the average similarity was taken from all languages for which data
181 microroles are readable because of overlap, but some major role clusters are
still discernible. Basically, what happens is that the average coexpression across
many languages identies clusters of microroles which approximately represent
one considers a cluster to be) (see Cysouw 2014 for a detailed discussion). To
the lein Figure 3 there is a cluster of agent-like roles, at the bottom there is a
cluster of patient-like roles, and at the top right there is a cluster of instruments.
that the spatial closeness of the microrole labels in Figure 3 was computed on the
microroles in Figure 2b was done manually.us, while the intermediate position
Identifying semantic role clusters and alignment types via microrole coexpression tendencies471
To the extent that cross-linguistic coexpression tendencies are due to semantic
resentation of the language-independent semantic similarities between the mi-
croroles. No subjective judgment of semantic similarity (e.g. between the ‘hidden
6.Mapping languages on the microrole map
the location of the 181 microroles was taken from the multidimensional scaling
with the squares being used for the agent-like microroles on the le-hand side, the
circles for patient-like microroles as well as goal-like, addressee-like, and recipi-
ent-like microroles, and the triangle being used for instrument-like microroles in
Identifying semantic role clusters and alignment types via microrole coexpression tendencies473
guages, with distribution lines to help us recognize the clusters of coding sets. We
coding set of each microrole is taken into account, both theag and the index.)
into discrete types, we quantify the pairwise similarity between all languages. In
this way we are able to investigate global similarities between languages (i.e. to de-
how similar each language is to each and every other language. We can then use
ties (e.g. a hierarchical clustering as shown in Figure 6 below).
larity from a single language was used.erefore, when two microroles used the
weighted Pearson correlation to establish the similarity between two language-
for the fact that the microroles are not equally distributed across all possible func-
like and patient-like microroles than instrument-like or goal-like roles. Any un-
alized in Figure 3).is means that pairs of microroles that are far apart are given
greater weight in the language comparison, and pairs of microroles that are close
archical clustering in Figure 6.13A few clusters of languages are indicated in this
Identifying semantic role clusters and alignment types via microrole coexpression tendencies475
In this paper, we have shown that it is possible to arrive at a role clustering
ging and indexing) for each microrole of 87 verbs in 25 languages from around
Figure 6.Hierarchical clustering of similarities in microrole coexpression (i.e. alignment
the world, allowing us to measure the similarities between the microroles and
thus arrive at mesorole-like clusters (Figure 3), and in a next step to show the
patient coded alike than monomorphemic verbs, cf. Malchukov 2013). Such similarities are not
Identifying semantic role clusters and alignment types via microrole coexpression tendencies477
more similar in Balinese than in Bora. All we are claiming is that they are treated more similarly
13.For this hierarchical clustering we used the functionhclustfrom the base package of the
statistical environment R. Specically, we used thecompletemethod for the clustering here.
Wälchli, Bernhard & Michael Cysouw. 2012. Lexical typology through similarity semantics:
Identifying semantic role clusters and alignment types via microrole coexpression tendencies479
Appendix 1:e 87 verb meanings and microroles
Identifying semantic role clusters and alignment types via microrole coexpression tendencies481
Identifying semantic role clusters and alignment types via microrole coexpression tendencies483
In this paper, I argue that “depth of analysis” does not deserve the prestige that it is sometimes given in general linguistics. While language description should certainly be as detailed as possible, general linguistics must rely on worldwide comparison of languages, and this cannot be based on language-particular analyses. Rigorous quantitative comparison requires uniform measurement, and this implies abstracting away from many language-particular peculiarities. I will illustrate this on the basis of ergative patterns, starting out from I.A. Mel’čuk’s (1981) proposal for Lezgian. This proposal was not successful, but why not? And why is Baker’s (2015) theory of dependent case likewise unsuccessful? By contrast, quantitative worldwide research has found striking similarities of ergative coding patterns, which can be explained by the efficiency theory of asymmetric coding. I will argue that this success is due to a more cautious approach to understanding Human Language, which does not rely on the Mendeleyevian vision for grammar (that all grammars are made from the same innate building blocks).
In this paper, we present an overview of the methods associated with semantic maps, focusing on current challenges and new avenues for research in this area, which are at the core of the contributions to this special issue. Among the fundamental questions are: (1) the validity of the basic assumption, namely, to what extent does coexpression reflect semantic similarity; (2) the central problem of identifying analytical primitives in the domain of semantics; (3) the methods of inference used for creating coexpression maps and the representation techniques (graph structure vs. Euclidean space) as well as their respective merits (including the goodness of fit of the models); and (4) the use of semantic maps to support diachronic and synchronic descriptions of individual languages. In order to illustrate and discuss key aspects, we conduct an experiment in the semantic field of emotions, for which we construct a classical semantic map based on the dataset of CLICS3.
We describe two mathematical representations for what have come to be called “semantic maps”, that is, representations of typological universals of linguistic co-expression with the aim of inferring similarity relations between concepts from those universals. The two mathematical representations are a graph structure and Euclidean space, the latter as inferred through multidimensional scaling. Graph structure representations come in two types. In both types, meanings are represented as vertices (nodes) and relations between meanings as edges (links). One representation is a pairwise co-expression graph, which represents all pairwise co-expression relations as edges in the graph; an example is CLICS. The other is a minimally connected co-expression graph – the “classic semantic map”. This represents only the edges necessary to maintain connectivity, that is, the principle that all the meanings expressed by a single form make up a connected subgraph of the whole graph. The Euclidean space represents meanings as points, and relations as Euclidean distance between points, in a specified number of spatial dimensions. We focus on the proper interpretation of both types of representations, algorithms for constructing the representations, measuring the goodness of fit of the representations to the data, and balancing goodness of fit with informativeness of the representation.
Instead of de!ning semantic roles on the basis of the interpretation of lexical predi-cates, I will show that it is possible to induce semantic roles from the usage of case-like markers across a wide variety of languages. The assumptions behind this pro-posal are, !rst, that semantic roles are strongly contextually determined and, sec-ond, that similarity in coding of contextual roles across many di""erent languages shows which contexts evoke the same (or better: very similar) semantic roles. This approach to the investigation of semantic roles will be exempli!ed by an investiga-tion of case-like marking in a parallel text across a sample of !fteen languages. On this basis, a semantic map of contextual roles can be established, and it will be shown that higher level abstractions, like semantic roles or even macro-roles, can be statistically derived from this diversity of marking across many languages. Further, a typology of alignment systems can be derived statistically.
By using this site, you consent to the processing of your personal data, the storing of cookies on your device, and the use of similar technologies for personalization, ads, analytics, etc. For more information or to opt out, see ourPrivacy Policy
",5.0