Professional Documents
Culture Documents
Textbook Non Canonically Case Marked Subjects The Reykjavik Eyjafjallajokull Papers Johanna Barddal Ebook All Chapter PDF
Textbook Non Canonically Case Marked Subjects The Reykjavik Eyjafjallajokull Papers Johanna Barddal Ebook All Chapter PDF
https://textbookfull.com/product/non-market-economies-in-the-
global-trading-system-the-special-case-of-china-james-j-
nedumpara/
https://textbookfull.com/product/perspectives-on-causation-
selected-papers-from-the-jerusalem-2017-workshop-elitzur-a-bar-
asher-siegal/
https://textbookfull.com/product/marked-02-0-marked-by-pain-1st-
edition-cece-rose/
https://textbookfull.com/product/incipient-the-marked-saga-6-1st-
edition-bianca-scardoni/
Calm Technology Principles and Patterns for Non
Intrusive Design 1st Edition Amber Case
https://textbookfull.com/product/calm-technology-principles-and-
patterns-for-non-intrusive-design-1st-edition-amber-case/
https://textbookfull.com/product/an-informed-faith-the-position-
papers-of-r-j-rushdoony-1st-edition-r-j-rushdoony/
https://textbookfull.com/product/social-determinants-of-health-
in-non-communicable-diseases-case-studies-from-japan-katsunori-
kondo/
https://textbookfull.com/product/midlife-and-magically-marked-
the-seventh-path-1-1st-edition-leona-knight/
https://textbookfull.com/product/advocacy-bar-
manuals-9-aug-2018-edition-the-city-law-school/
Studies in Language Companion Series (SLCS)
ISSN 0165-7763
Founding Editor
Werner Abraham
University of Vienna I University of Munich
Editors
Werner Abraham Elly van Gelderen
University of Vienna I University of Munich Arizona State University
Editorial Board
Bernard Comrie Elisabeth Leiss
University of California, Santa Barbara University of Munich
William Croft Marianne Mithun
University of New Mexico University of California, Santa Barbara
Osten Dahl Heiko Narrog
University of Stockholm Tohuku University
Gerrit J. Dimmendaal Johanna L. Wood
University of Cologne University of Aarhus
Ekkehard Konig Debra Ziegeler
Free University of Berlin University of Paris III
Christian Lehmann
University of Erfurt
Volume200
Non-Canonically Case-Marked Subjects. The Reykjavik-Eyjafjallajokull papers
Edited by Johanna Barodal, Na'ama Pat-Eland Stephen Mark Carey
Non-Canonically
Case-Marked Subjects
The Reykjavik-Eyj afj allaj okull papers
Edited by
J6hanna Barodal
Ghent University
Na'ama Pat-El
The University of Texas, Austin
DOI 10.1075/Slcs.200
Cataloging-in-Publication Data available from Library of Congress:
LCCN 2018027446 (PRINT) /2018043505 (E-BOOK)
ISBN 978 90 272 0147 8 (HB)
ISBN 978 90 272 6351 3 (E-BOOK)
CHAPTER 1
Introduction: The Reykjavik-Eyjafjallajokull papers 1
]ohanna Barodal
CHAPTER 2
Non-nominative and depersonalized subjects in the Balkans:
Areality vs. genealogy 23
Victor A. Friedman and Brian D. ]oseph
CHAPTER3
Affective constructions in Tsezic languages 55
Bernard Comrie, Diana Porker and Zaira Khalilova
CHAPTER 4
A macrorole approach to dative subjects
Patrick Farrell and Beatriz Willgohs
CHAPTER 5
Dative case and oblique subjects 115
Robert D. Van Valin, fr.
CHAPTER 6
Word order as a subject test in Old Icelandic 135
Johannes G. /onsson
VI Non-canonically Case-marked Subjects
CHAPTER?
The diachrony of non-canonical subjects in Northwest Semitic 155
Na'ama Pat-E[
CHAPTER 8
Case marking of predicative possession in Vedic: The genitive,
the dative, the locative 181
Serena Danesi and ]6hanna Barodal
CHAPTER 9
Accusative sickness? A brief epidemic in the history of German 213
Tonya Kim Dewey and Stephen Mark Carey
Afterword
CHAPTER 10
Forty years in the search of a/the subject 241
Andrej Malchukov
CHAPTER 11
What is a subject: The nature and validity of subject tests 257
]6hanna Barodal and 7h6rhallur Eyth6rsson
Introduction
The Reykjavik-Eyjafjallajokull papers
J6hanna Barodal
Ghent University
1. Background
https://doi.org/Io.1075/slcs.zoo.mbar
© 2018 John Benjarnins Publishing Company
2 J6hanna Barddal
de Hoop & de Swart 2008; Barodal & Chelliah 2009; Malchukov & Spencer 2009;
Donohue & Barodal20 11; Cennamo, Barodal & Gelderen 20 12; Gelderen, Cennamo
& Barodal20 13; Sedant & Kulikov 20 13; and Helasvuo & Huumo 20 15). Early work
on Modern Icelandic was carried out within Lexical Functional Grammar (Zaenen,
Mating & Thniinsson 1985) and Government and Binding Theory (Sigurosson
1989, 1991). Subjecthood and subject properties have also been dealt with in Role
and Reference Grammar (Van Valin & LaPolla 1997), Functional Grammar (Dik
1997), Basic Linguistic Theory (Dixon 1997), and Radical Construction Grammar
(Croft 2001; Barodal2006), inter alia.
What all the approaches mentioned above have in common is their emphasis on
the distributional and behavioral properties of subjects, like their ability to be left un-
expressed in conjoined clauses and control infinitives, their ability to figure in object
and subject raising, and to control reflexives, as well as their word order distribution
in different types of clause structures. However, as research among non-Indo-Eu-
ropean languages started progressing and the structure of more languages across
different language families became known, it also became evident that the behavior
of syntactic subjects across languages is quite heterogeneous, and the spread of sub-
ject properties across languages turned out to be more variegated than expected,
especially given the assumed existence of universal grammar, with corresponding
problems accommodating this variegated behavior within such a grammar.
Thus, in more recent years, there has been a change in focus from universal
properties of subjects to language-specific properties of subjects (Dryer 1997; Croft
2001; Culicover & Jackendoff 2005; Van Valin 2005; Barodal2006; Bickel2011).
It has become doubtful as to whether or not a universal concept of subject can be
maintained because of significant typological differences between languages and
the great variety of constructions relevant to this concept. For example, ergative
languages code subjects in two separate ways depending on the transitivity of the
verb (Kikusawa 2002; Queixlos & Gildea 2010). And even in such closely related
languages like Icelandic and German, less than one third of the assumed universal
subject properties, as discussed by Keenan (1976), have been regarded as valid tests
for both languages (Barodal2006: 55; Barodal, EythOrsson & Dewey 20 14; Barodal
& EythOrsson this volume). Additional subject tests have even been identified for
some languages, like Jonsson (1996) who suggests as many as 27 tests for Modern
Icelandic, consisting of syntactic behaviors and distribution that clearly distinguish
between subjects and objects.
Unrelated linguistic frameworks have also integrated the concept of subject
in different ways into their formalism. In the Minimalist Framework, the subject
is defined through the matching of a bundle of interpretable phi- or EPP-features,
while in classical Government and Binding Theory (GB) and related theories, sub-
ject properties are attributed to a particular functional/structural position in the
Chapter 1. Introduction 3
and predicates denoting purely relational and ontological states. Oblique subjects
may also denote possession, modality and evidentiality, as well as featuring the
intransitive variant of causative pairs (anticausatives} in some Indo-European lan-
guages (e.g. Sandal2011; Barodal2014; Bjarnadottir 2014; Cennamo, Eythorsson
& Barodal2015}. In a wider typological perspective, it remains to be established
which semantic features are language-family-specific and which are generally found
cross-linguistically.
3· Summaries
instead (here labelled R for Receptors). This revised RRG approach links semantic
representations with syntactic structure, while at the same time allowing for both
universal mapping and language-specific mapping.
Van Valin's chapter deals with dative arguments, comparing data from differ-
ent constructions in languages like Icelandic, German, English, Spanish, French,
Basque, Japanese, Bolivian Quechua, Kannada, and Dyirbal. The constructions rele-
vant for the present study are oblique subject constructions, ergative constructions,
impersonal reflexive constructions, dative object constructions, and ditransitive
constructions. Van Valin argues, on the basis of this comparison, that the dative
is neither grammatically nor semantically motivated; that is, the dative is not con-
sistently used to mark a grammatical function or a semantic role. He then brings
additional data from causative constructions into the discussion, arguing that case
alternations for the causee argument demonstrate the lack of any inherent mean-
ing of the dative. Rather, Van Valin argues that the dative is a default case and not
semantically-motivated lexical case, and hence that it is found in structures where
rules assigning other cases do not apply or are blocked from applying.
Jonsson, in his contribution, focuses on the role of word order distribution
for identifying grammatical relations in Old Icelandic, an early Indo-European
language which is known for having freer word order than its successor, Modern
Icelandic. The major difference between the two language stages, relevant for gram-
matical relations, is that Old Icelandic allows OV word order within the verb phrase,
much like Modern German where OV is obligatory, while Modern Icelandic does
not. Preverbal position is a problematic subject test in Old Icelandic since objects
may be topicalized; postverbal position is also problematic since heavy and indefi-
nite subjects tend to occur to the right in the clause. This leaves only the middle field
and the subsequent task of differentiating between subjects and objects when both
arguments follow the finite verb, a well-known problem in research on other early
North Germanic languages, for instance, Old Swedish. However, due to pronomi-
nal object scrambling in Old Icelandic, a word order in which pronominal objects
precede full NP subjects, J6nsson is able to conclude that word order, involving
pronominal subjects and objects following a finite verb, can indeed be used to dis-
tinguish between subjects and objects in Old Icelandic. Not surprisingly, it turns
out that oblique subject-like arguments show the same word order distribution in
this respect as nominative subjects.
Pat-El, in her chapter, demonstrates that constructions with oblique subject-like
arguments are found in earlier periods of two Semitic languages, namely Biblical
Hebrew and Aramaic, in addition to Modern Hebrew. Both these corpus languages
belong to the North-West Semitic branch, hence the data do not favor a recon-
struction of this pattern for Proto-Semitic. With regard to syntactic behavior, these
oblique subject-like arguments behave as nominative subjects in most respects;
Chapter 1. Introduction 7
they occupy the subject slot in neutral word order, they are always definite, they
are obligatorily marked with prepositions, a property not shared with objects or
adverbials, and they control subject omission in absolutives. On these grounds,
Pat-El argues that oblique subject-like arguments should be analyzed as syntactic
subjects and that an object analysis is excluded. It is postulated that the category
of oblique subject has developed from free benefactives, which were reanalyzed as
obligatory core arguments in only two relevant Semitic languages. This analysis is
in part based on the distribution of oblique subjects vs. free datives, as some other
Semitic branches indeed have free datives but no oblique subjects, and in part on
a stipulation that infinitive subjects have been reanalyzed as complements in the
relevant North-West Semitic languages, due to the high topicality of free datives.
Danesi and Bar<Sdal zoom in on predicative possession in one of the earliest
attested stages oflndo-European, namely Vedic Sanskrit, confining the data and the
analysis to the oldest Vedic text, the Rigveda. Three predicative constructions exist
all expressing possession, one with the possessor marked in the genitive, a second
with the possessor marked in the dative and a third one with the possessor marked
in the locative case. The low frequency of the locative construction suggests a rather
marginal status in Vedic. At the level of information structure, the genitive and the
dative constructions also show considerable differences. The genitive construction
expresses a possessee-oriented relation, with the possessee being topical while the
possessor represents focal information. The dative construction, in contrast, has a
tendency to be possessor oriented and is neutral with regard to information struc-
ture. Another difference between the two is that the genitive construction expresses
concrete possession, while the dative construction expresses more often than not
abstract possession. This very fact speaks against claims in the literature that the
dative possessive construction gave rise to the dative experiencer construction
through a development from concrete possession like 'book' to abstract "posses-
sion'' like 'happiness', 'sorrow', 'sadness'.
Dewey and Carey bring to the fore an undocumented change in subject case
marking that took place in Old and Middle High German. This change is manifested
in a set of five verbal predicates, which must be reconstructed as selecting for dative
subjects, based on the comparative evidence. During the Old High German period,
however, these predicates occur systematically with accusative subjects instead of
the expected dative. These are in githahti koman 'begin to think', leid sin 'be sad',
niot sin 'desire', oth sin 'desire' and wuntar sin 'wonder', which are all compositional
predicates constructed with the verb 'be' and a noun or an adjective. In addition,
there are further six predicates, which must also be reconstructed with a dative, that
show variation between accusative and dative subject marking in Old and Middle
High German; these are angusten 'fear', gilimphan 'find appropriate', gerinnan 'lack',
thunken 'think, seem', (un)willon 'feel nausea' and wola sin 'have it good'. This change
8 J6hanna Barodal
from dative to accusative subjects is rare in the Germanic languages, while the op-
posite change, from accusative to dative subjects, is well known, especially from
Icelandic, under the label Dative Sickness or Dative Substitution. Dewey & Carey
argue that the change from dative to accusative is pragmatically motivated in that
the accusative signals a higher degree of affectedness than the dative.
The volume concludes with an extensive afterword, containing sections
by Malchukov, on the one hand, and Barodal and Eyth6rsson, on the other. In
Malchukov's contribution an overview is given of the quest for a universal subject
within the typological literature, where the nature of the subject in ergative lan-
guages and their behavior distinct from the subject in accusative languages must
be taken into account. It is argued that further reasons for splitting the subject up
come from the observation that a selection of the subject properties seems to derive
from topicality, while others derive from agentivity. Splits in subject behavior are
not only found in ergative languages, but several accusative languages also show
splits, for instance between nominative and oblique subjects. Such splits in subject
behavior have been dealt with differently in various theoretical frameworks, includ-
ing splits observed in diachronic research on subject status. Malchukov proposes,
for instance, that oblique subjects may arise through a path where transitive clauses
develop into intransitive ones, via a process that he labels "transimpersonal sce-
nario': in which a generic or indefinite subject is gradually dropped, resulting in a
reanalysis of the argument structure, including the former transitive object becom-
ing the new intransitive subject. A further split of the subject tests, aligning with
accusative vs. ergative languages, respectively, is also discussed in that imperatives,
control, reflexivization and conjunction reduction are typical subject tests for accu-
sative languages, while nominalization, resultatives, verbal plurality and possessor
ascension are typical subject tests for ergative languages. Finally, Malchukov sets
out a program for how it may be possible to complete the quest for subject, that
proponents of typology started some 40 years ago.
The section by Malchukov is then juxtaposed to a section by Baradal and
Eythorsson who point to some of the major analytical problems arising from the
fact that most current scholarship on potential subject status of oblique subject-like
arguments does not employ an independent definition of subject. Instead, this work
contains analyses based on the subject tests alone, which then function as a stand-in
definition for subject, in the absence of a more principled methodology. This is
particularly problematic when an oblique subject-like argument passes some of
the subject tests, but not others, which in turn requires the relevant scholar to
start measuring the subject tests against each other in order to choose between
them, again often on unprincipled grounds. This results in a gradient view of the
subject concept, yielding "one-fourth': "one-third" or "half" a subject. With an in-
dependently applied subject definition, in contrast, an argument is either a subject
Chapter 1. Introduction 9
or not, irrespective of the number of subject tests in a given language. The analysts
new task is then to unearth any potential restrictions on the constructions used as
subject tests, as well as a possible incompatibility between that construction and
predicates selecting for oblique subject-like arguments in the relevant language.
One eminent example is German, and the infamous Icelandic-German dichot-
omy which assumes that oblique subject-like arguments are syntactic subjects in
Icelandic, but syntactic objects in German. A further study of the subject tests in
Icelandic and German which yield different results between the two languages all
turn out to involve ellipsis. Hence, instead of analyzing potential oblique subjects in
German as objects, a more adequate generalization can be made, namely that there
are firmer restrictions on subject ellipsis in German than in Icelandic.
There is no doubt that the distribution of oblique subjects is areal. There are, for
instance, great structural similarities between the dative subject construction in the
In do-European languages in Europe, on the one hand, and Basque in the South and
Kartvelian in the North-East, on the other. At present, little or no research has been
carried out comparing the structure and the semantics of these constructions across
these language families. However, areal facts of this type raise the issue of contact vs.
inheritance. Comparative In do-European research has shown that it is indubitable
that the dative subject construction is inherited in the In do-European languages
(cf. Barodal et al. 2012). This, however, does not exclude contact in and of itself.
In recent research on the early Germanic languages, Barodal et al. (2016)
suggest a computational method aiming at teasing apart contact and inheritance.
Barodal et al. (2016) are unable to document any differences between Old Norse
and Old and Middle English, on the one hand, and the remaining Germanic lan-
guages on the other, despite the early contact between Germanic tribes occupying
these areas. Similarly, no evidence is found in the data for contact between Old
Swedish, on the one hand, and Continental Germanic, on the other, in spite of
the well-known heavy low German influence on Old Danish and Old Swedish
during the high Middle Ages. However, this computational method is based on
a comparison between narrowly-defined semantic classes. Given that no effect of
contact can be documented through this comparison of geographical areas which
we know have been in contact, it is reasonable to assume that any borrowings
between the Germanic languages during these periods must have involved items
belonging to the same narrowly-defined semantic classes of verbs. Other methods
must then be used to distinguish between inheritance and contact for oblique sub-
ject constructions.
10 Johanna Bari'ldal
1. For arguments against Luraghi's analysis, see Danesi, Johnson & Bari'ldal (2017).
Chapter 1. Introduction 11
Mithun 2008; Barodal & Eyth6rsson 2009; Sedant 2013; Malchukov this volume),
as few languages provide the historical depth needed to uncover such develop-
ments. For the In do-European languages, for instance, which are recorded as far
back as 4,000 years, dative subject constructions are reconstructable for Proto-Indo-
European (Barodal & Smitherman 2013; Barodal et al. 2013; Danesi, Johnson &
Barodal 20 17), pushing their origin further back into pre-documented history.
Another assumed origin of oblique subjects within Indo-European linguistics
postulates that they have developed from the mihi-est possessive construction
(cf. Bauer 2002; Fedriani 2011). This development is laid out by Fedriani (2011)
who observes that the only difference between the possessive construction and
the experiencer construction is that the nominative object in the possessive con-
struction is concrete, say a book, while the corresponding nominative object in the
experiencer construction is abstract, denoting say happiness, sorrow, sadness, etc.
Hence, the development must have been from concrete possessees to abstract pos-
sessees, Fedriani claims. Put differently, through an extension of the nominative
slot, from concrete to abstract, a new construction arises, namely the experiencer
construction, which is structurally identical to the dative possessive construc-
tion, with the only difference between the two being the degree of concreteness/
abstractness of the nominative possessee. Danesi and Barodal, however, show in
their chapter that such a development is unlikely, as already in the earliest Vedic
texts, the mihi-est possessive construction occurs with abstract possessees, while
concrete possessees usually occur in the predicative possessive construction of
the type "The car is John's:' That is, already in the oldest Indo-European texts, the
mihi-est construction shows a clear tendency to be instantiated by abstract nouns,
hence no development from concrete to abstract can be documented in the history
of the In do-European languages.
Dewey & Carey's investigation is a pertinent contribution to the ever-growing
body of research on differential case marking, both subject and object case marking
(Bossong 1985; Aissen 2003; Abraham 2006; de Hoop & Swart 2008; Malchukov
2008; Donohue & Barodal2011; Heusinger & Kaiser 2011, Malchukov & de Hoop
2011, Iemmolo & Klumpp 2014). It is well known from Modern Icelandic that var-
iation in subject case marking does not reflect any differences in meaning (J6nsson
& Eyth6rsson 2005; Barodal2009, 2011), but can rather be taken as evidence for a
functional merger between the categories of accusative and dative subjects. Dewey
and Carey show that the variation they document in Old and Middle High German,
i.e. from dative subjects to accusative subjects, is motivated by pragmatic factors,
signifying a higher degree of affectedness (cf. Tsunoda 1985) for the accusative
than the dative.
Van Valin's investigation is a contribution to the theoretical debate on the
role of dative case marking, be it structural, lexical, semantic, abstract, or default
(Svenonious 2002; Sigurosson 2003; Woolford 2006; Barodal 2011), based on its
12 J6hanna Barlldal
Our hope with this volume is to contribute to one general and one specific en-
terprise: The first enterprise is the ongoing search and disclosure of the facts of
non-canonically case-marked subjects within and across languages, be these lexical,
semantic or syntactic. The second and more specific enterprise is to provide a forum
for discussing how these facts should be interpreted and analyzed.
A pressing issue in the field of noncanonical coding of arguments is the
variation found with regard to the subject tests in languages like, for instance,
Lithuanian, Russian, Polish and Spanish, where potential oblique subjects only pass
a limited number of the subject tests, but not necessarily all. Such facts have tradi-
tionally been used to argue either that the potential oblique subject is not a subject
at all, but an object (Zaenen, Maling & Thniinsson 1985; M0rck 1994; Faarlund
2001) or "a special subcategory of" indirect object (Moore & Perlmutter 2000), or
more recently to argue for the existence of a category labelled "semi-subject" or
"quasi-subject" (Farrell2005; Conti 2008; Holvoet 2013; Serzant 2013). The problem
with the earlier analyses is that no evidence for object status is presented, in practice
Chapter 1. Introduction 13
turning the object concept into a waste-paper basket category, while the problem
with the later analyses is that they introduce into the grammar a category of uncer-
tain and ambiguous nature, with no real status in the grammar of these languages.
Several questions arise: Why is it a given that potential oblique subjects should
not be analyzed as syntactic subjects if they only pass a portion of the subject tests?
Why is a subject analysis of potential oblique subjects called to a higher standard
than a subject analysis of ordinary nominative subjects? Why are corresponding
object analyses of potential oblique subjects not subjected to any standard at all? Is
there any gain in inventing labels, and corresponding categories, like "semi -subject':
"pseudo-subject" or "quasi subject"? Are some of the so-called subject tests per-
haps case tests, targeting arguments with the same morphological coding, etc. We
believe that the present volume contributes to some of these issues, including con-
tributions to some of the most current debates in the literature on non-canonically
case-marked subjects in the world's languages.
We envision more research carried out on languages where potential oblique
subjects do not pass all the subject tests, namely languages like, for instance,
Lithuanian, Russian, Polish and Spanish, in order to identify the relevant restric-
tions on the constructions that constitute the tests. It is important to uncover
whether these are case restrictions, or perhaps restrictions that somehow only allow
for default-marked arguments, like the nominative. Clearly, several subject tests in
ergative languages may be of this type, thus excluding subjects of transitive verbs
and intransitive verbs from showing the same syntactic behavior across languages,
resulting, among others, in the well-known concept of syntactic ergativity.
A second strand of research that is important for the issue of subject status
in corpus languages is research on word order in the so-called "free word order"
languages. Such research must aim for disentangling the different pragmatic con-
ditions associated with the different word order patterns in a given language. It has
often been argued, implicitly and explicitly, that freedom in word order excludes the
use of word order distribution as a subject test (Reis 1982; Faarlund 2001). This is
simply not true, as long as it can be identified how information is packaged within
a language, word order may be used as a viable subject test (cf. Le Mair et al. 20 17).
In Barc5dal (2006), for instance, I identified the difference between neutral word
order and topicalizations in ordinary declarative clauses in Modern German on
the basis of intonation. This in turn made it possible to argue for a subject analysis
for German "oblique subject" constructions, as opposed to a "topicalized object
analysis" (cf. also Barc5dal, Eyth6rsson & Dewey 2014).
Recognizing how information structure affects word order distribution is not
only of validity for languages spoken currently but also for historical languages,
like Latin and Ancient Greek, not to mention Sanskrit. Only through such work
can it be established which of the relevant word orders conveys neutrality with
14 J6hanna Barddal
Acknowledgements
For considerably raising the quality of this volume, we would like to thank the following inter-
nal, external and editorial reviewers: Werner Abraham, Peter Arkadiev, Alessandra Caviglia,
Concepci6n Company, Serena Danesi, Tonya Kim Dewey-Findell, Jan Terje FaarluQd, Stephanie
Harves, John Huehnergard, J6hannes Gisli J6nsson, Martin Kiimmel, Santeri Palviainen, Yael
Reshel, Misumi Sadler, Wolfgang Schulze, Peter Sells, Ilja A. Sedant, Graeme Trousdale, Kevin
J. Tuite, Robert Van Valin, Carlotta Viti, and Tamar Zewi. We thank the series editors, Elly van
Gelderen and Werner Abraham, for having welcomed this volume into their book series, and the
acquisition editor, Kees Vaes, for his support.
Chapter 1. Introduction 15
References
Abraham, Werner. 2006. Bare and prepositional differential case marking: The exotic case of
German (and Icelandic) among all of Germanic. In Kulikov, Malchukov & de Swart (eds),
115-145.
Aikhenvald, Alexandra Y., Dixon, Robert M. W. & Onishi, Masayuki (eds). 2001. Non-Canonical
Marking of Subjects and Objects [Typological Studies in Language 46]. Amsterdam: John
Benjamins. https://doi.orgho.1075/tsl.46
Aissen, Judith. 2003. Differential object marking: Iconicity vs. economy. Natural Language &
Linguistic Theory 21: 435-448. https://doi.orgho.1023/A:1024109008573
Andrews, Avery D. 1976. The VP complement analysis in Modern Icelandic. North Eastern
Linguistic Society 6: 1-21.
Bar(ldal, J6hanna. 2000. The subject is nominative! On obsolete axioms and their deep-
rootedness. In 17th Scandinavian Conference of Linguistics, Carl-Erik Lindberg & Steffen
Nordahl Lund (eds), 93-117. Odense: Institute of Language and Communication.
Bar(ldal, J6hanna. 2004. The semantics of the impersonal construction in Icelandic, German and
Faroese: Beyond thematic roles. In Focus on Germanic Typology, Werner Abraham (ed.),
105-137. Berlin: Akademie Verlag.
Bar(ldal, J6hanna. 2006. Construction-specific properties of syntactic subjects in Icelandic and
German. Cognitive Linguistics 17(1): 39-106. https://doi.orgho.1515/COG.2oo6.oo2
Bar(ldal, J6hanna. 2009. The development of case in Germanic. In Bar(ldal & Chelliah (eds),
123-159.
Bar(ldal, J6hanna. 2011. The rise of dative substitution in the history of Icelandic: A diachronic
construction grammar approach. Lingua 121(1): 60-79.
https://doi.orgho.1o16/j.lingua.2o1o.o7.oo7
Bar(ldal, J6hanna. 2014. Syntax and syntactic reconstruction. In The Routledge Handbook of
Historical Linguistics, Claire Bowern & Bethwyn Evans (eds), 343-373. London: Routledge.
Bar(ldal, J6hanna, Arnett, Carlee, Carey, Stephen Mark, Eyth6rsson, Th6rhallur, Jenset, Gard
B., Kroonen, Guus & Oberlin, Adam. 2016. Dative subjects in Germanic: A computational
analysis of lexical semantic verb classes across time and space. STUF: Language Typology
and Universals 69(1): 49-84.
16 J6hanna Barlldal
Barlldal, J6hanna, Bjarnad6ttir, Valgerllur, Danesi, Serena, Dewey, Tonya Kim, Eyth6rsson,
Th6rhallur, Fedriani, Chiara & Smitherman, Thomas. 2013. The story of 'Woe'. Journal ofIndo-
European Studies 41{3-4): 321-377.
Barlldal, J6hanna & Chelliah, Shobhana L. (eds). 2009. The Role of Semantic, Pragmatic and
Discourse Factors in the Development of Case [Studies in Language Companion Series 108].
Amsterdam: John Benjamins. https://doi.org/1o.1o7s/slcs.1o8
Barlldal, J6hanna & Eyth6rsson, Th6rhallur. 2009. The origin of the oblique subject construc-
tion: An Indo-European comparison. In Grammatical Change in Indo-European Languages
[Current Issues in Linguistic Theory 305], Vit Bubenik, John Hewson & Sarah Rose (eds),
179-193. Amsterdam: John Benjamins. https://doi.org/1o.1o7s/cilt.305.19bar
Barlldal, J6hanna, Eyth6rsson, Th6rhallur & Dewey, Tonya Kim. 2014. Alternating predicates
in Icelandic and German: A sign-based construction grammar account. Working Papers in
Scandinavian Syntax 93: 50-101.
Barlldal, J6hanna & Smitherman, Thomas. 2013. The quest for cognates: A reconstruction of
oblique subject constructions in Proto-Indo-European. Language Dynamics and Change
3(1): 28-67.
Barlldal, J6hanna, Smitherman, Thomas, Bjarnad6ttir, Valgerllur, Danesi, Serena, Jenset, Gard B.
& McGillivray, Barbara. 2012. Reconstructing constructional semantics: The dative subject
construction in Old Norse-Icelandic, Latin, Ancient Greek, Old Russian and Old Lithua-
nian. Studies in Language 36(3): 511-547. https://doi.org/1o.1o7s/sl.36.3.03bar
Bhaskararao, Peri & Subbarao, Karumuri Venkata (eds). 2004. Non-Nominative Subjects, Vols.
1-11 [Typological Studies in Language 60-61]. Amsterdam: John Benjamins.
Bickel, Balthasar. 2004. The syntax of experiencers in the Himalayas. In Bhaskararao & Subbarao,
Vol. 1, 77-111.
Bickel, Balthasar. 2011. Grammatical relations typology. In The Oxford Handbook of Language
Typology, Jae Jung Song (ed.), 399-444. Oxford: OUP.
Bjarnad6ttir, Valgerllur. 2014. The oblique anticausative in Lithuanian. Baltistica 49(1): 15-39.
https://doi.org/1o.1S388/baltistica.49.1.2211.
Cennamo, Michela, Barlldal, J6hanna & van Gelderen, Elly (eds). 2012. Variation and change in
argument realization. A special guest-edited issue in Transactions of the Philological Society
110(3): 311-442. https://doi.org/1o.1111/j.1467-968X.2o12.01323.x
Cennamo, Michela, Eyth6rsson, Th6rhallur & Barlldal, J6hanna. 2015. Semantic and (morpho)
syntactic constraints on anticausativization: Evidence from Latin and Old Norse-Icelandic.
Linguistics 53{4): 677-729. https://doi.org/1o.151SIIing-2o1s-oo15
Comrie, Bernard. 1973. The ergative: Variations on a theme. Lingua 32: 239-253.
https://doi.org/1o.1o16/oo24-3841(73)90044-2
Conti, Luz. 2008. Synchronie und Diachronie des altgriechischen Genitivs als Semisubjekt.
Historische Sprachforschung 121: 94-113.
Croft, William. 2001. Radical Construction Grammar: Syntactic Theory in Typological Perspective.
Oxford: OUP. https://doi.org/1o.1093/acprof:oso/g780198299554.001.0001
Culicover, Peter W. & Jackendoff, Ray. 2005. Simpler Syntax. Oxford: OUP.
https://doi.org/1o.1093/acprof:oso/g780199271092.001.oo01
Danesi, Serena, Johnson, Cynthia A. & Barlldal, J6hanna. 2017. Between the historical languages
and the reconstructed language: An alternative approach to the Gerundive + "Dative of
Agent" construction in Indo-European. Indogermanische Forschungen 122: 143-188.
Danesi, Serena, Cynthia A. Johnson & J6hanna Bar<ldal. 2018. Where does the modality of
Ancient Greek modal verbs come from? The relation between modality and oblique case
marking. Journal of Greek Linguistics 18(1): 45-92.
Chapter 1. Introduction 17
Dik, Simon C. 1997. The Theory of Functional Grammar, Part 1: The Structure of the Clause, 2nd
edn. Berlin: Mouton de Gruyter.
Dixon, Robert M. W. 1997. The Rise and Fall of Languages. Cambridge: CUP.
https://doi.orgl1o.1017/CB097B051161206o
Donohue, Cathryn & Badldal, J6hanna (eds). 2011. Empirical approaches to morphological case.
A special guest-edited issue in Morphology 21(3-4): 481-654.
https://doi.org/1o.1oo7/511525-011-9187-4
Dryer, Martin. 2006. Descriptive theories, explanatory theories, and basic linguistic theory. In
Catching Language: The Standard Challenge of Grammar Writing, Felix Ameka, Alan Charles
Dench & Nicholas Evans (eds), 207-234. Berlin: Mouton de Gruyter.
Dryer, Matthew S. 1997. Are grammatical relations universal? In Essays on Language Function
and Language Type. Dedicated to T. Giv6n, Joan Bybee, John Haiman & Sandra A.
Thompson (eds), 117-143. Amsterdam: John Benjamins. https://doi.orgl1o.1o7s/z.82.ogdry
Eyth6rsson, Th6rhallur & Bar(ldal, J6hanna. 2005. Oblique subjects: A common Germanic in-
heritance. Language 81(4): 824-881. https://doi.orgl1o.13531lan.2oos.o173
Farrell, Patrick. 2005. Grammatical Relations. Oxford: OUP.
Faarlund, Jan Terje. 2001. The notion of oblique subject and its status in the history of Icelandic.
In Grammatical Relations in Change [Studies in Language Companion Series 56], Jan Terje
Faarlund (ed.), 99-135. Amsterdam: John Benjamins. https://doi.orgl1o.1o7s/slcs.s6.osfaa
Fedriani, Chiara. 2011. Experiential metaphors in Latin: Feelings were containers, movements
and things possessed. Transactions of the Philological Society 109(3): 307-326.
https://doi.orgl1o.1111/j.1467-968X.2011.01284.x
Fedriani, Chiara. 2014. Experiential Constructions in Latin. Leiden: Brill.
https://doi.orgl1o.1163/9789004257832
van Gelderen, Elly, Cennamo, Michela & Bar(ldal, J6hanna (eds). 2013. Argument Structure in
Flux: The Naples-Capri Papers [Studies in Language Companion Series 131]. Amsterdam:
John Benjamins. https://doi.orgl1o.1o7s/slcs.131
Greenberg, Joseph. 1963. Some universals of grammar with particular reference to the order
of meaningful elements. In Universals of Language, Joseph H. Greenberg (ed.), 73-113.
Cambridge MA: The MIT Press.
Hansen, Bjorn. 2016. Beyond b-subjects: Testing subjecthood in Croatian modal constructions.
Talk presented at the workshop "Forty Years after Keenan 1976: Subject Properties and
Subject Tests'; 7-9 September, Ghent.
Haspelmath, Martin. 2004. Does linguistic explanation presuppose linguistic description?
Studies in Language 28: 554-579. https://doi.orgl1o.107s/sl.28.3.o6has
Helasvuo, Marja-Liisa & Huumo, Tuomas (eds). 2015. Subjects in Constructions- Canonical and
Non-Canonical [Constructional Approaches to Language 16]. Amsterdam: John Benjamins.
Hermon, Gabriella. 1985. Syntactic Modularity. Dordrecht: Foris.
von Heusinger, Klaus & Kaiser, Georg A. 2011. Affectedness and differential object marking in
Spanish. Morphology 21 (3-4): 593-617. https://doi.org/1o.1007/S1152S-010-91n-y
Hock, Hans H. 1990. Oblique subjects in Sanskrit? In Verma & Mohanan (eds), 119-139.
Hock, Hans Heinrich & Bashir, Elena. 20 16. The Languages and Linguistics of South Asia: A
Comprehensive Guide. Berlin: De Gruyter Mouton.
Holvoet, Axe!. 2013. Obliqueness, quasi-subjects and transitivity in Baltic and Slavonic. In
Sedant & Kulikov (eds), 257-282.
de Hoop, Helen & de Swart, Peter (eds). 2008. Differential Subject Marking. Dordrecht: Springer.
Iemmolo, Giorgio & Klumpp, Gerson (eds). 2014. Differential object marking: Theoretical and
empirical issues. A special issue of Linguistics 52(2): 271-602.
18 J6hanna Bari'!dal
J6nsson, J6hannes Gisli. 1996. Clausal Architecture and Case in Icelandic. PhD dissertation,
University of Massachusetts, Amherst.
J6nsson, J6hannes G. & Eyth6rsson, Th6rhallur. 2005. Variation in subject case marking in insu-
lar Scandinavian. Nordic Journal of Linguistics 28(2): 223-245.
https://doi.org/10.1017/So3325B6sosoo1435
Keenan, Edward L. 1976. Towards a universal definition of subject. In Li (ed.), 303-333.
Kikusawa, Ritsuko. 2002. Proto Central Pacific Ergativity: Its Reconstruction and Development in
the Fijian, Rotuman and Polynesian languages. Canberra: Pacific Linguistics.
Kulikov, Leonid, Malchukov, Andrej & de Swart, Peter (eds). 2006. Case, Valency and Transitivity
[Studies in Language Companion Series 77]. Amsterdam: John Benjamins.
https://doi.orgho.1o7s/slcs.n
Landau, Idan. 2009. The Locative Syntax of Experiencers. Cambridge MA: The MIT Press.
Le Mair, Esther, Johnson, Cynthia A., Frotscher, Michael, Eyth6rsson, Th6rhallur & Bari'!dal,
J6hanna. 2017. Position as a behavioral property of subjects: The case of Old Irish. Indo-
germanische Forschungen 122: 111-142.
Li, Charles N. (ed.). 1976. Subject and Topic. New York NY: Academic press.
Luraghi, Silvia. 2016. The dative of agent in Indo-European languages. STUF- Language
Typology and Universals 69(1): 15-47.
Malchukov, Andrej L. 2008. Animacy and asymmetries in differential case marking. Lingua
118(2): 203-221. https://doi.orgho.1o16/j.lingua.2007.02.oos
Malchukov, Andrej L. & de Hoop, Helen. 2011. Tense, aspect, and mood based differential case
marking. Lingua 121(1): 35-47. https://doi.orgho.1o16/j.lingua.2010.07.oo3
Malchukov, Andrej L. & Spencer, Andrew (eds). 2009. The Oxford Handbook of Case. Oxford:
OUP.
Masica, Colin P. 1976. Defining a Linguistic Area: South Asia. Chicago IL: University of Chicago
Press.
Mithun, Marianne. 2008. The emergence of agentive systems. In The Typology of Semantic
Alignment Systems, Mark Donohue & S0ren Wichmann (eds), 297-333. Oxford: OUP.
https://doi.orgho.1093/acprof:oso/978019923838S.003.0012
Moore, John & Perlmutter, David M. 2000. What does it take to be a dative subject? Natural
Language and Linguistic Theory 18: 373-416. https://doi.orgho.1023/A:1oo6451714195
M0rck, Endre. 1994. The distribution of subject properties and the acquisition of subjecthood
in the West Scandinavian languages. In Language Change and Language Structure: Older
Germanic Languages in a Comparative Perspective, Toril Swan, Endre M0rck & Olaf J.
Westwik (eds), 159-194. Berlin: Mouton de Gruyter.
https://doi.orgho.1515/97B3110886573.159
Onishi, Masayuki. 2001. Introduction: Non-canonically marked subjects and objects; parameters
and properties. In Aikhenvald, Dixon & Onishi (eds), 1-51.
Primus, Beatrice. 2011. Case marking typology. In The Oxford Handbook of Linguistic Typology,
Jae Jung Song (ed.), 303-321. Oxford: OUP.
Queixal6s, Francesc & Gildea, Spike. 2010. Manifestations of Ergativity in Amazonia. In Ergativ-
ity in Amazonia [Typological Studies in Language 89], Spike Gildea & Francesc Queixal6s
(eds), 1-25. Amsterdam: John Benjamins. https://doi.orgho.107s/tsi.B9.01int
Rakosi, Gyorgy. 2006. Dative Experiencer Predicates in Hungarian. Utrecht: LOT.
Reis, Marga. 1982. Zum Subjektbegriff im Deutschen. In Satzglieder im Deutschen. Vorschliige
zur syntaktischen, semantischen und pragmatischen Fundierung, Werner Abraham (ed.),
171-211. Tiibingen: Gunter Narr.
Chapter 1. Introduction 19
Sandal, Catrine. 2011. Akkusative subjekt og antikausativitet i nommt (Accusative Subjects and
Anticausativity in Old Norse). MA thesis, University of Bergen.
Sasse, Hans-Jiirgen. 1978. Subjekt und Ergativ: Zur pragmatischen Grundlage primarer gramma-
tischer Relationen. Folia Linguistica 12:219-252. https://doi.orgho.1515/flin.1978.12.3-4.219
Sedant, Ilja. 2013. The diachronic typology of non-canonical subjects. In Serzant & Kulikov
(eds), 313-360.
Sedant, Ilja A. & Kulikov, Leonid (eds). 2013. The Diachronic Typology ofNon-Canonical Subjects
[Studies in Language Companion Series 140]. Amsterdam. John Benjamins.
https://doi.orgho.1o7s/slcs.140
Shibatani, Masayoshi. 1999. Dative subject constructions twenty-two years later. Studies in the
Linguistic Sciences 29(2): 45-76.
Sigur(lsson, Halld6r Armann. 1989. Verbal Syntax and Case in Icelandic. PhD dissertation, Lund
University.
Sigur(lsson, Halld6r Armann. 1991. Icelandic case-marked PRO and the licensing oflexical ar-
guments. Natural Language and Linguistic Theory 9: 327-362.
https://doi.orgho.1007/BFoo134679
Sigur(lsson, Halld6r A. 2003. Case: Abstract vs. morphological. In New Perspectives on Case
Theory, Ellen Brandner & Heike Zinsmeister (eds), 223-268. Stanford CA: CSLI.
Steever, Sanford B. (ed.) 1998. The Dravidian Languages. London: Routledge.
Svenonius, Peter. 2002. Icelandic case and the structure of events. The Journal of Comparative
Germanic Linguistics 5(1-3): 197-225. https://doi.orgho.1023/A:1021252206904
Tsunoda, Tasaku. 1985. Remarks on transitivity. Journal of Linguistics 21(2): 385-396.
https://doi.orgho.1017/Soo22226700010318
Van Valin Jr., Robert D. 1991. Another look at Icelandic case marking and grammatical relations.
Natural Language and Linguistic Theory 9: 145-194. https://doi.orgho.1007/BFoo133328
Van Valin Jr., Robert D. 2005. Exploring the Syntax-Semantics Interface. Cambridge: CUP.
https://doi.org/10.1017/CB09780S11610S78
Van Valin Jr., Robert D. 2009. Case in role and reference grammar. In The Handbook of Case,
Andrej L. Malchukov & Andrew Spencer (eds), 102-120. Oxford: OUP.
Van Valin Jr., Robert D. & LaPolla, Randy J. 1997. Syntax: Structure, Meaning and Function.
Cambridge: CUP. https://doi.orgho.1017/CB09781139166799
Verma, Mahendra K. & Mohanan, Karravanur Puthanvettil (eds). 1990. Experiencer Subjects in
South Asian Languages. Stanford CA: CSLI.
Woolford, Ellen. 2006. Lexical case, inherent case, and argument structure. Linguistic Inquiry
37(1): 111-130. https://doi.orgho.1162/oo2438906775321175
Yoon, James H. 2004. Non-nominative (major) subjects and case stacking in Korean. In
Bhaskararao & Subbarao, Vol. 2, 265-314.
Zaenen, Annie, Maling, Joan & Thnii.nsson, Hoskuldur. 1985. Case and grammatical functions:
The Icelandic passive. Natural Language and Linguistic Theory 3:441-483.
https://doi.orgl1o.1 oo7/BFoo133285
PART I
The languages of the Balkan sprachbund are surveyed here with regard to their
constructions that show non-nominative subjects, typically in impersonal con-
structions. The issue of origins is considered, specifically as to whether these
constructions represent inheritances from some earlier stage of the relevant !an-
guages or instead reflect the effects of contact. In the end, it is argued that a mix
of areality, i.e. contact, and genealogy, i.e. inheritance, is needed to explain these
constructions, with a nod required as well to typologically common patternings.
1. Introduction 1
Masica (1976, 2001) claims that the construction he calls a 'dative subject experi-
encer' is an areal characteristic of South Asia and the Caucasus, and notes further
1. Throughout, in providing glosses for our examples, we follow the Leipzig Glossing Rules
(LGR}, though we augment the Leipzig list with: "IMPF" for imperfect (past tense}; "IPFV" for im-
perfective (aspect}; "MP" for medio-passive, in regard to certain nonactive verb forms in Albanian
and Greek; "PC" for the Albanian so-called particle of concord; and "SP" for the subordinating par-
ticle found in all the Balkan languages. We also deviate from the LGR in the Balkan Romance ex-
amples (Aromanian, Daco-Romanian, and Megleno-Romanian) in using a hyphen after the weak
object pronouns not to indicate a morpheme boundary but to follow the conventional orthography
for those languages; also, "VOC" here refers to a vocative particle, not a vocative per se. For the
element with the form se in Balkan Slavic and Balkan Romance and pe in Romani, we variably
give the gloss "INTR'' for intransitive or "REFr' for reflexive, or "MP~ depending on the particular
usage in a given example. We use the following abbreviations for languages discussed here: Alb,
for Albanian; Armn, for Aromanian; BCSM, for Bosnian-Croatian-Serbian-Montenegrin; Big, for
Bulgarian; Grk, for Greek; Mac, for Macedonian; MR, for Meglenoromanian; Rmn, for Romanian;
Rmi, for Romani; and, Trk, for Turkish; note also BR, for Balkan Romance, taking in Armn, MR,
and Rmn; and BS, for Balkan Slavic, taking in Big and Mac.
https://doi.org/to.t075/slcs.2oo.o2fri
© 2018 John Benjarnins Publishing Company
24 Victor A. Friedman and Brian D. Joseph
that it can be found to some extent in Europe too. Although subject-type has re-
ceived considerable attention for individual branches oflndo-European as such (e.g.
Barodal et al. 2012), those subsequently experiencing contact in the Balkans (see
Map 1) - Albanian, Greek, In die, Romance, and Slavic - have not been examined
Maramure~
Transylvania
ROMANIA
Wallachia....,.......... GAGAUZ
BuLGARIAN
JUOEZMO
MEDITERRANEAN
SEA
Romani
Balkan
I South
Turkic
IINorth Turkish Gagauz
Vlax Rumelian
South East
North West
Hellenic Balkan Slavic
Balkan Romance Greek Torlak (SE BCS)
Albanian Timok-NiSava
Geg Tosk Daco-Romanian Vlah North
Wallachia Prizren-South Morava
Lab Aromanian South
Moldavia Macedonian
<;:am North/West Tsakonian
Banat East
Arvanitika South/Eest West
Crijana Judezmo
Arberesh Maramure~
Megleno-Romanlan Bulgarian
East
lstro-Romanian Tsarnareca East
West
Other West
SUB-CLASS I.—ENTOMOSTRACA.
Order I. Branchiopoda.[14]
Sub-Order 1. Phyllopoda.
Associated with the development of the carapace, in this and in the next
family, is a remarkable condition of the lateral eyes, which are sessile on
the dorsal surface of the head, and near the middle line, the median eye
being slightly in front of them. During embryonic life a fold of skin grows
over all three eyes, so that a chamber is formed over them, which
communicates with the exterior by a small pore in front.
In the Limnadiidae the body is laterally compressed, and the carapace
is so large that at least the post-cephalic part of the body, and generally
the head also, can be enclosed within it.
In Limnetis (Fig. 3) the dorsal
surface of the head is bent
downwards and is much
compressed, the carapace being
attached to it only for a short
distance near the dorsal middle line.
The sides of the carapace are bent
downwards, and their margins can
be pulled together by a transverse
adductor muscle, so that the whole
structure forms an ovoid or
spheroidal case, from which the
head projects in front, while the rest
Fig. 3.—Limnetis brachyura, × 15. (After
of the body is entirely contained G. O. Sars.)
within it. When the adductor muscle
is relaxed the edges of the carapace
gape slightly, like the valves of a Lamellibranch shell, and food-particles
are drawn through the opening thus formed into the ventral groove by the
movements of the thoracic feet, locomotion being chiefly effected by the
rowing action of the second antennae, as in the Cladocera, to which all the
Limnadiidae present strong resemblances in their method of locomotion,
in the condition of the carapace, and in the form of the telson.
In Limnadia and Estheria the carapace projects not only backwards
from the point of attachment to the head, but also forwards, so that the
head can be enclosed by it, together with the rest of the body.
In all these genera the carapace is flexible along the middle dorsal line;
in Estheria especially the softening of the dorsal cuticle goes so far that a
definite hinge-line is formed, and this, together with the deposition of the
lateral cuticle in lines concentrically arranged round a projecting umbo,
gives the carapace a strong superficial likeness to a Lamellibranch shell,
for which it is said to be frequently mistaken by collectors.
The eyes of the Limnadiidae are enclosed in a chamber formed by a
growth of skin over them, as in Apodidae, but the pore by which this
chamber communicates with the exterior is even more minute than in
Apus. The paired eyes are so close together that they may touch
(Limnadia, Estheria) or fuse (Limnetis); they are farther back than in the
Apodidae, while the ventral curvature of the head causes the median eye
to lie below them. In all these points the eyes of the Limnadiidae are
intermediate between those of Apus and those of the Cladocera.
Dorsal Organ.—A structure very characteristic of adult Phyllopods is
the “dorsal organ” (Figs. 2, 5, D.O), whose function is in many cases
obscure. It is always a patch of modified cephalic ectoderm, supplied by a
nerve from the anterior ventral lobe of the brain on each side; but its
characters, and apparent function, differ in different forms. In the
Branchipodidae the dorsal organ is a circular patch, far forward on the
surface of the head (Figs. 2, 5, D.O). Its cells are arranged in groups,
which remind one of the retinulae in a compound eye; each cell contains a
solid concretion, and the concretions of a group may be so placed as to
look like a badly-formed rhabdom. Claus,[16] who first called attention to
this structure in the Branchipodidae, regarded it as a sense-organ. In
Apodidae the dorsal organ is an oval patch of columnar ectoderm,
immediately behind the eyes; it is slightly raised above the surrounding
skin, and is covered by a very delicate cuticle (with an opening to the
exterior?), and below it is a mass of connective tissue permeated by blood;
Bernard has suggested that it is an excretory organ.
Most Limnadiidae resemble the Cladocera in the possession of a
“dorsal organ” quite distinct from the above; in Limnetis and Estheria it
has the form of a small pit, lined by an apparently glandular ectoderm,
and this is its condition in many Cladocera; in Limnadia lenticularis it is
a patch of glandular epithelium on a raised papilla. Limnadia has been
observed to anchor itself to foreign objects by pressing its dorsal organ
against them, and many Cladocera do the same thing; Sida crystallina,
for example, will remain for hours attached by its dorsal organ to a
waterweed or to the side of an aquarium. Structures resembling a dorsal
organ occur in the larvae of many other Crustacea, but the presence of
this organ in the adult is confined to Branchiopods, and indeed in many
Cladocera it disappears before maturity. It is certain that the sensory and
adhesive types of dorsal organ are not homologous, especially as
rudimentary sense-organs may exist on the head of Cladocera together
with the adhesive organ.
The telson differs considerably in the different genera. In the
Branchipodidae[17] the anus opens directly backwards; and the telson
carries two flattened backwardly directed plates, one on each side of the
anus, the margins of each plate being fringed with plumose setae. In
Artemia the anal plates are rarely as large as in Branchipus, and never
have their margins completely fringed with setae; in A. salina from
Western Europe, and in A. fertilis (Fig. 4, A) from the Great Salt Lake of
Utah, there is a variable number of setae round the apical half of each
lobe, but in specimens of A. salina from Western Siberia the number of
setae may be very small, or they may be absent; in the closely allied A.
urmiana from Persia the anal lobes are well developed in the male, each
lobe bearing a single terminal hair, but they are altogether absent in the
female. Schmankewitch and Bateson have shown that there is a certain
relation between the salinity of the water in which Artemia salina occurs
and the condition of the anal lobes, specimens from denser waters having
on the whole fewer setae; the relation is, however, evidently very complex,
and further evidence is wanted before any more definite statements can
be made.
Fig. 4.—A, Ventral view of the anal region in Artemia fertilis, from the
Great Salt Lake; B, ventral view of the telson and neighbouring parts of
Lepidurus productus; C, side view of the telson and left anal lobe of
Estheria (sp.?).
In the Apodidae the anal lobes have the form of two-jointed cirri, often
of considerable length; in Apus the anus is terminal, but in Lepidurus
(Fig. 4, B) the dorsal part of the telson is prolonged backwards, so as to
form a plate, on the ventral face of which the anus opens, much as in the
Malacostraca.
In the Limnadiidae (Fig. 4, C) the telson is laterally compressed and
produced, on each side of the anus, into a flattened, upwardly curved
process, sharply pointed posteriorly, and often serrate; the anal lobes are
represented by two stout curved spines, while in place of the dorsal
prolongation of Lepidurus we find two long plumose setae above the
anus. In the characters of the telson and anal lobes, as in those of the
head, the Limnadiidae approximate to the Cladocera. In Limnetis
brachyura the ventral face of the telson is produced into a plate
projecting backwards below the anus, in a manner which has no exact
parallel among other Crustacea.
The appendages of the Phyllopoda are fairly uniform in character,
except those affected by the sexual dimorphism, which is usually great.
Fig. 5.—Chirocephalus diaphanus, male. Side view of head, showing
the large second antenna, A2, with its appendage Ap, above which is
seen the filiform first antenna; D.O, dorsal organ; E1, median eye.
Of the cephalic appendages, the first antennae are generally small, and
are never biramous; in Branchipus and its allies they are simple unjointed
rods, in some species of Artemia they are three-jointed, in Apus they are
feebly divided into two joints, while in Estheria they are many-jointed.
The second antennae are the principal organs of locomotion in the
Limnadiidae, where they are large and biramous; in all other Phyllopoda
they are uniramous in the female, being either unjointed triangular plates
as in Chirocephalus (Fig. 2), or minute vestigial filaments as in Apus, in
which genus Zaddach, Huxley, and Claus have all failed to find any trace
of a second antenna in some females. In the male Branchipodidae the
second antennae are modified to form claspers, by which the female is
seized, the various degrees of complication which these claspers exhibit
affording convenient generic characters. In Branchinecta each second
antenna is a thick, three-jointed rod, the last joint forming a claw, while
the second joint is serrate on its inner margin; in Branchipus the base is
much thickened, and bears on its inner side a large filament (perhaps
represented by the proximal tubercle of Branchinecta and Artemia),
which looks like an extra antenna. In Streptocephalus the terminal joint
of the antenna is bifid, and there is a basal filament like that of
Branchipus; in Chirocephalus
diaphanus (Figs. 5, 6) the main
branch of the antenna consists of
two large joints, the terminal joint
being a strong claw with a serrated
process at its base, while the
proximal joint bears two
appendages on its inner side; one of
these is a small, subconical tubercle,
the second is more complicated,
consisting of a main stem and five
outgrowths. The main stem is
many-jointed and flexible, its basal
joint being longer than the others,
and bearing on its outer side a large,
triangular, membranous appendage,
and four soft cylindrical
appendages, the main stem and its
appendages being beset with
curious tubercles, ending in short
spines, whose structure is not
understood. Except during the act of
copulation this remarkable Fig. 6.—Chirocephalus diaphanus.
Second antenna of male, uncoiled.
apparatus is coiled on the inner side
of the antennary claw, the jointed
stem being so coiled that it is often compared to the coiled proboscis of a
butterfly, and the triangular membrane folded like a fan beside it, so that
much of the organ is concealed, and the general appearance of the head is
that shown in Fig. 5. During copulation, the whole structure is widely
extended.
The males of Artemia (Fig. 7) have the second antenna two-jointed, the
basal joint bearing an inner tubercle, the terminal joint being flattened
and bluntly pointed, its outer margin provided with a membranous
outgrowth. In A. fertilis the breadth of the second joint varies greatly, the
narrower forms presenting a certain remote resemblance to
Branchinecta. In the males of Polyartemia the second antennae have a
remarkable branched form not easily comparable with that found in other
Branchipodidae.
The cephalic jaws are fairly uniform throughout the order. The
mandibles have an undivided molar surface, and no palp; the first maxilla
is very generally a triangular plate, with a setose biting edge; mandibles
and maxillae are covered by the
labrum. The second maxilla
generally lies outside the chamber
formed by the labrum, and is a
simple oval plate, with or without a
special process for the duct of the
kidney.
The thoracic limbs, in front of the
genital segments, are not as a rule
differentiated into anterior
Fig. 7.—Artemia fertilis. Front view of the maxillipedes and posterior
head of a male, showing the large second locomotive appendages, as in higher
antennae, A.2; A.1, first antennae. forms; we have seen, however, that
all these limbs take part in the
prehension of food, and except in
the Limnadiidae they all assist in locomotion. One of the middle thoracic
legs of Artemia (Fig. 8, A) has a flattened stem, with seven processes on
its inner, and two on its outer margin. The gnathobase (gn) is large, and
fringed with long plumose setae, each of which is jointed; this is followed
by four smaller “endites” (or processes on the median side), and then by
two larger ones, the terminal endite (the sixth, excluding the gnathobase)
being very mobile and attached to the main stem by a definite joint. On
the outer side are two processes; a proximal “bract,” a flat plate with
crenate edges, partly divided by a constriction into two, and a distal
process, cylindrical and vascular, called by Sars and others the
“epipodite.” In other Branchipodidae we have essentially the same
condition, except that the fifth endite often becomes much larger than in
Artemia, throwing the terminal endite well over to the outer edge of the
limb; such a shift as this, continued farther, might well lead to the
condition found in the Limnadiidae, or Apodidae, where the lobe which
seems to represent the terminal endite of Artemia is entirely on the outer
border of the limb, forming what most writers have called the exopodite
(Lankester’s “flabellum”).[18] In the two last-named families the basal
exite or bract of the Branchipodidae does not appear to be represented.
Fig. 8.—A, Thoracic limb of Chirocephalus diaphanus; B, prehensile
thoracic limb of male Estheria. gn, Gnathobase; 1–6, the more distal
endites.
The limbs of the Apodidae are remarkable in two ways; those in front of
the genital opening (very constantly ten pairs) are not so nearly alike as in
most genera of the sub-order, the first two pairs especially having the axis
definitely jointed, while the endites are elongated and antenniform;
further, while the first eleven segments bear each a single pair of limbs, as
is usual among Crustacea, many of the post-genital segments bear several
pairs; thus in Apus cancriformis there are thirty-two post-cephalic
segments in front of the telson, the first eleven having each one pair of
limbs, while the next seventeen have fifty-two pairs between them, the
last four segments having none.
In all the Phyllopoda some of the post-cephalic limbs are modified for
reproductive purposes; in the Branchipodidae the last two pairs (the 12th
and 13th generally, the 20th and 21st in Polyartemia) are so modified in
both sexes. In the female these appendages fuse at an early period of
larval life, and surround the median opening of the generative duct (Fig.
2); in the male the two pairs also fuse, but traces of the limbs are left as
eversible processes round the paired openings of the vasa deferentia.
In the other families, one or more limbs of the female are adapted for
carrying or supporting the eggs. In the Apodidae the appendages of the
eleventh segment have the exopodite in the form of a rounded,
watchglass-shaped plate, fitting over a similarly shaped process of the
axis of the limb, so that a lens-shaped box is formed, into which the eggs
pass from the oviduct. In Limnadiidae the eggs are carried in masses
between the body and the carapace, and are kept in position by special
elongations of the exopodites of two or three legs, either those near the
middle of the thorax (Estheria, Limnadia), or at its posterior end
(Limnetis). In female Limnetis the last thoracic segments bear two
remarkable lateral plates, which apparently also help to support the eggs.
In the male Limnadiidae, the first (Limnetis) or the first two thoracic feet
(Limnadia, Estheria) are prehensile (Fig. 8, B).
Alimentary Canal.—The mouth of the Phyllopoda is overhung by the
large labrum, so that a kind of atrium is formed, outside the mouth itself,
in which mastication is performed; numerous unicellular glands, opening
on the oral face of the labrum, pour their secretion into the atrial
chamber, and may be called salivary, though the nature of their secretion
is not known. The mouth has commonly two swollen and setose lips,
running longitudinally forwards from the bases of the first maxillae, and
often wrapping round the blades of the mandibles. It leads into a vertical
oesophagus, which opens into a small globular stomach, lying entirely
within the head; the terminal part of the oesophagus is slightly
invaginated into the stomach, so that a valvular ring is formed at the
junction of the two. The stomach opens widely behind into a straight
intestine, which runs backwards to about the level of the telson, where it
joins a short rectum, leading to the terminal or ventral anus. The stomach
and intestine are lined by a columnar epithelium, and covered by a thin
network of circularly arranged muscle-fibres; the rectum has a flatter
epithelium, and radial muscles pass from it to the body-wall, so that it can
be dilated. The only special digestive glands are two branched glandular
tubes, situated entirely within the head, which open into the stomach by
large ducts, one on each side. In Chirocephalus the gastric glands are
fairly small and simple; in the Apodidae their branches are more complex
and form a considerable mass, filling all that portion of the head which is
not occupied by the nervous system and the muscles. Backwardly directed
gastric glands, like those of the higher Crustacea, are not found in
Branchiopods; both forms occur together in the genus Nebalia, but with
this exception the forwardly-directed glands are peculiar to
Branchiopods.
Heart.—In Branchipus and its allies, and in Artemia, the heart
extends from the first thoracic segment to the penultimate segment of the
body, and is provided with eighteen pairs of lateral openings, one pair in
every segment through which it passes except the last; it is widely open at
its hinder end, and is prolonged in front for a short distance as a cephalic
aorta, the rest of the blood-spaces being lacunar.
In most, at least, of the other Branchiopods, the heart is closed behind
and is shortened; in Apus and Lepidurus it only extends through the first
eleven post-cephalic segments, while in the Limnadiidae it is shorter still,
the heart of Limnetis passing through four segments only. In all cases
there is a pair of lateral openings in every segment traversed by the heart.
The blood of the Branchipodidae and Apodidae contains dissolved
haemoglobin, the quantity present being so small as to give but a faint
colour to the blood in Branchipus, while Artemia has rather more, and
the blood of Apus is very red. The only other Crustacea in which the blood
contains haemoglobin are the Copepods of the genus Lernanthropus,[19]
so that the appearance of this substance is as irregular and inexplicable in
Crustacea as in Chaetopods and Molluscs.
The nervous system of Branchipus may be described as an
illustration of the condition prevailing in the group. The brain consists of
two closely united ganglia, in each of which three main regions may be
distinguished; a ventral anterior lobe, a dorsal anterior lobe, and a
posterior lobe. The ventral anterior lobes give off nerves to the median
eye, to the dorsal organ, and to a pair of curious sense-organs,
comparable with the larval sense-knobs of many higher forms, situated
one on each side of the median eye; in late larvae Claus describes the
terminal apparatus of each frontal sense-organ as a single large
hypodermic cell; W. K. Spencer[20] has lately described several terminal
cells, containing peculiar chitinous bodies, in the adult. The homologous
sense-organs of Limnetis are apparently olfactory. The dorsal anterior
lobes give off the large nerves to the lateral eyes, while the posterior lobes
supply the first antennae. The oesophageal connectives have a coating of
ganglion-cells, and some of these form the ganglion of the second
antenna, the nerve to this appendage leaving the connective just behind
the brain. The post-oral nerve-cords are widely separate, each of them
dilating into a ganglion opposite every appendage, the two ganglia being
connected by two transverse commissures. The ganglia of the three
cephalic jaws, so often fused in the higher Crustacea, are here perfectly
distinct. Closely connected with each thoracic ganglion is a remarkable
unicellular gland, opening to the exterior near the middle ventral line; it
is conceivable that these cells may be properly compared with the larval
nephridia of a Chaetopod,[21] but no evidence in support of such a
comparison has yet been adduced.
Behind the genital segments, where there are no limbs, the nerve-cords
run backwards without dilating into segmental ganglia, except in the
anterior two abdominal segments where small ganglionic enlargements
occur. In Apodidae, on the other hand, those segments which carry more
than one pair of appendages have as many pairs of ganglia, united by
transverse commissures, as they have limbs.
A stomatogastric nervous system exists in Apus, where a nerve arises
on each side from the first post-oral commissure, and runs forward to join
its fellow of the opposite side on the anterior wall of the oesophagus.
From the loop so formed a larger median and a series of smaller lateral
nerves pass to the wall of the alimentary canal. A second nerve to the
oesophagus is given off from the mandibular ganglion of each side.
Reproductive Organs.—In Chirocephalus the ovaries (Fig. 2, Ov)
are hollow epithelial tubes, lying one on each side of the alimentary canal,
and extending from the sixth abdominal segment forwards to the level of
the genital opening; at this point the two ovaries are continuous with
ducts, which bend sharply downwards and open into the single uterus
contained within the projecting egg-pouch and opening to the exterior at
the apex of that organ. Short diverticula of the walls of the uterus receive
the ducts of groups of unicellular glands, the bodies of which contain a
peculiar opaque secretion, said to form the eggshells. In Apodidae the
ovaries are similar in structure, but they are much larger and branch in a
complex manner, while each ovary opens to the exterior independently of
the other in the eleventh post-cephalic segment; nothing like the median
uterus of the Branchipodidae being formed. The epithelium of the ovarian
tubes proliferates, and groups of cells are formed; one becoming an ovum,
the others being nutrient cells like those which will be more fully
described in the Cladocera.
In Chirocephalus the testes are tubes similar in shape and position to
the ovaries, each communicating in front with a short vas deferens, which
dilates into a vesicula seminalis on its way to the eversible penis; an
essentially similar arrangement is found in all Branchipodidae, but in
Apodidae and Limnadiidae there is no penis.
All the Branchiopoda are dioecious,[22] and many are parthenogenetic.
Among Branchipodidae Artemia is the only genus known to be
parthenogenetic, but parthenogenesis is common in all Apodidae, while
the males of several species of Limnadia are still unknown, although the
females are sometimes exceedingly common. In Artemia, generations in
which the males are about as numerous as the females seem to alternate
fairly quickly with others which contain only parthenogenetic females; in
Apus males are rarely abundant, and often absent for long periods; during
five consecutive years von Siebold failed to discover a male in a locality in
Bavaria, though he examined many thousands of individuals; near
Breslau he found on one occasion about 11 per cent of males (114 in 1026),
but in a subsequent year he found less than 1 per cent; the greatest
recorded percentage of males is that observed by Lubbock in 1863, when
he found 33 males among 72 individuals taken near Rouen.
The eggs of most genera can resist prolonged periods of desiccation,
and indeed it seems necessary for the development of many species that
the eggs should be first dried and afterwards placed in water. Many eggs
(e.g. of Chirocephalus diaphanus and Branchipus stagnalis) float when
placed in water after desiccation, the development taking place at the
surface of the water.
Habitat.—All the Phyllopoda, except Artemia, are confined to
stagnant shallow waters, especially to such ponds as are formed during
spring rains, and dry up during the summer. In waters of this kind the
species of Branchipus, Apus, etc., develop rapidly, and produce great
numbers of eggs, which are left in the dried mud at the bottom after
evaporation of the water, where they remain quiescent until a fresh rainy
season. The mud from the beds of such temporary pools often contains
large numbers of eggs, which may be carried by wind, on the legs of birds,
and by other means, to considerable distances. Many exotic species have
been made known to European naturalists by their power of hatching out
when mud brought home by travellers is placed in water. The water of
stagnant pools quickly dissolves a certain quantity of solid matter from
the soil, and often receives dissolved solids through surface drainage from
the neighbouring land; such salts may remain as the water evaporates, so
that the water which remains after evaporation has proceeded for some
time may be very sensibly denser than that in which the Branchiopods
were hatched; these creatures must therefore be able to endure a
considerable increase in the salinity of the surrounding waters during the
course of their lives. My friend Mr. W. W. Fisher points out that the
plants present in such a pond would often precipitate the carbonate of
lime, so that this might be removed as evaporation went on, but that
chlorides would probably remain in solution; from analyses which Mr.
Fisher has been kind enough to make for me, it is seen that this happened
in a small aquarium in my laboratory, in which Chirocephalus diaphanus
lived for four months. In April, mud from the dry bed of a pond, known to
contain eggs of Chirocephalus, was placed in this aquarium in Oxford,
and water was added from the tap. Oxford tap-water contains about 0·3
grm. salts per litre, the chlorine being equivalent to 0·023 grm. NaCl.
Water was added from time to time during May and June, but in July
evaporation was allowed to proceed unchecked. At the end of July there
was about half the original volume of water, the Chirocephalus being still
active; the residue contained 0·96 grm. dissolved solids per litre, with
chlorine equal to 0·19 grm. NaCl, so that the percentage of chlorides was
about eight times the initial percentage, but there were only three and a
fifth times the original amount of total solid matter in solution, the
carbonate of lime having precipitated as a visible film.
Some species of Branchipus (e.g. B. spinosus, M. Edw.) and of Estheria
(E. macgillivrayi, Baird, E. gubernator, Klutzinger) occur in salt pools,
but Artemia flourishes in waters beside whose salinity that endured by
any other Branchiopod is insignificant. In the South of Europe, Artemia
salina may be found in swarms, as it used to be found in Dorsetshire, in
the shallow brine-pans from which salt is commercially prepared; Rathke
quotes an analysis showing that a pool in the Crimea contained living
Artemia when the salts in solution were 271 grms. per litre, and the water
was said to have the colour and consistency of beer.
The behaviour of the animals in the water differs a little; in normal
feeding all the species swim with the back downwards, as has already
been said; the Branchipodidae rarely settle on the ground, or on foreign
objects, but the Apodidae occasionally wriggle along the bottom on their
ventral surface, and Estheria burrows in mud.
The greater number of species are found in pools in flat, low-lying
regions, and many appear to be especially abundant near the sea; Apus
cancriformis has, however, been found in Armenia at 10,000 feet above
sea level.
Wells and underground waters do not generally contain Phyllopods;
but a species of Branchipus and one of Limnetis, both blind, have been
described from the caves of Carniola.
One of the many puzzles presented by these creatures is the erratic way
in which they are scattered through the regions they inhabit; a single
small pond, a few yards or less in diameter, may be the only place within
many miles in which a given species can be found; in this pond it may,
however, appear regularly season after season for some time, and then
suddenly vanish.
Geographically, the Phyllopoda are cosmopolitan, representatives of
every family and of some genera (e.g. Streptocephalus, Lepidurus,
Estheria) being found in every one of the great zoological regions, though
a few aberrant genera are of limited range, thus Polyartemia is known
only from the northern Palaearctic and Nearctic regions,
Thamnocephalus only from the Central United States. The genus Artemia
is not at present known in Australia.[23] The only recorded British species
are Chirocephalus diaphanus, Artemia salina, and Apus cancriformis,[24]
but other continental islands, for example the West Indian group, are
better supplied. The distribution of the species is very imperfectly known,
but on the whole every main zoological region seems to have its own
peculiar species, which do not pass beyond its boundaries. Branchinecta
paludosa and Lepidurus glacialis are circumpolar, both occurring in
Norway, in Lapland, in Greenland, and in Arctic North America; but with
these exceptions the Palaearctic and Nearctic species seem to be distinct.
The European species Apus cancriformis occurs in Algiers, but the
relations between the species of Northern Africa as a whole and those of
Southern Europe on the one hand, or of Central and Southern Africa on
the other, have yet to be worked out.
The soft-bodied Branchipodidae are not known in the fossil condition;
[25]
an Apus, closely related to the modern A. cancriformis, has been
found in the Trias, but the most numerous remains have been left, as
might be expected, by the hard-shelled Limnadiidae; carapaces, closely
resembling those of the modern Estheria, are known in beds of all ages
from the Devonian period to recent times; these carapaces are in several
cases associated with fossils of an apparently marine type. None of the
fossil species differ in any important characters from those now living, so
that the Phyllopoda have existed in practically their present form for an
enormously long period; this fact, and the evidence that species of
existing genera were at one time marine, explain the wide distribution of
animals at present restricted to a remarkably limited range of
environmental conditions.