Johan Blomberg*
The expression of non-actual motion
in Swedish, French and Thai
DOI 10.1515/cog-2015-0025
Received March 9, 2015; revised July 31, 2015; accepted August 19, 2015
Abstract: Dynamic descriptions of static spatial situations, such as the road goes
through the forest have attracted a lot of attention across different semantic
theories. Analyses in terms of fictive motion and subjective motion have pro-
posed that such expressions are strongly motivated by universal cognitive and
conceptual factors. I present theoretical arguments for the conflation of several
different motivations in the literature. Instead of a single general motivation,
three distinct experiential motivations are presented under the term non-actual
motion. These experiential motivations are used to design an elicitation tool for
investigating non-actual motion cross-linguistically. Elicited descriptions from
speakers of Swedish, French and Thai suggest that such descriptions are con-
ventionalized in all three languages, which supports the universal character of
non-actual motion across languages. However, in expressing non-actual motion,
the language-specific resources for expressing actual motion are used.
Keywords: non-actual motion, fictive motion, subjective motion, motion
semantics, semantic typology
1 Introduction
The use of motion verbs to describe static situations, as in the sentences (1)–(4),
occupies a pivotal role in several semantic theories. Langacker (1990, 2006)
gives such sentences a central position in bridging the divide between the
concrete and the abstract in language and cognition. In the same examples,
Talmy sees a general “cognitive bias towards dynamism”(Talmy 2000a: 171–172)
whereas Matlock (2004a: 1390) claims that “language is structured the way it is
because of our natural ability to simulate motion”. Thus, it is claimed that these
sentences reveal how language is grounded in universal cognitive factors. From
this point of view it can then be conjectured that all languages would have the
resources to describe static situations with motion verbs.
Corresponding author: Johan Blomberg, Centre for Languages and Literature, Lund University,
Sweden & Institut für Sprache und Kommunikation, Technische Universität Berlin, Germany,
E-mail: [email protected]
Cognitive Linguistics 2015; 26(4): 657–696
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(1) The mountain range goes all the way from Mexico to Canada.(Talmy
2000a: 104)
(2) The mountain range goes all the way from Canada to Mexico. (Talmy
2000a: 104)
(3) The highway crawls through the city. (Matlock 2004b: 232)
(4) There is like this snaking road up the hills. (Brandt 2009: 583)
It is notable that several different terms have been used to denote the sentences
in (1)–(4). In reviewing previous literature, Blomberg and Zlatev (2014) found the
phenomenon explicated differently across various analyses. The previously
suggested terms virtual motion (Talmy 1983), fictive motion (Talmy 2000a),
abstract motion (Matlock 2010) and subjective motion (Langacker 1990; Matlock
2004a, 2004b; Matsumoto 1996) each have a different scope of meaning but are
nevertheless often used interchangeably. In order to capture a feature common
to previous analyses and pinpoint that the sentences superimpose a motion not
actually there, we proposed the more neutral term non-actual motion (NAM),
which will be used throughout this article.
[Non-actual motion] refer[s] to certain dynamic qualities of intentional acts that can be
seen as motivating the use of sentences with motion semantics to denote static situations.
[…] Thus, when we speak of non-actual motion sentences, we use this as a cover term for
all sentences in which (minimally) a motion verb is used to denote a situation that lacks
observed motion. (Blomberg and Zlatev 2014: 397)
In line with this inclusive definition, we argue that NAM-sentences are motivated
by three different experiences: (1) the capacity for self-motion connected to affor-
dances for motion (i.e., the perception of action potentials in the environment,
which entails an immediate connection between visual perception and action), (2)
visual scanning and (3) the imagination of movement. Based on these motivations,
a stimuli set for eliciting NAM-sentences was designed. With a material clearly
demarcating between several possible motivations, the production of NAM-
sentences in relation to properties of the situation could be investigated.
Given the importance attributed to such expressions, it is surprising to note
how scarce systematic cross-linguistic comparisons of this tendency have been.
Among the few publications with a comparative element are Matsumoto (1996),
Amagawa (1997), Rojo and Valenzuela (2004, 2009), and Stosic and Sarda
(2009). These studies have all found that motion verbs are used in describing
static configurations, but with varying constraints across languages. The present
658 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
study extends beyond the approach of previous studies in two ways. First, the
experiment was designed to differentiate between several possible motivating
factors. Second, the linguistic variation is related to different strategies for
expressing physical motion.
Swedish, French and Thai were selected due to their status in motion
typology as prime candidates of the three different types: s(atellite)-framed,
v(erb)-framed and equipollently-framed languages, respectively. Talmy’sfamous
binary motion event typology predicts that languages would express Path in the
main verb root (V-framing) or in a complement standing in what Talmy (2000b:
102) describes as a “sister relation the main verb root”(S-framing). This leaves
Manner (and additional information pertaining to the Figure’s movement, such as
Cause) to be either optionally expressed in V-framing languages (in, for example,
a gerund) or expressed by the main verb root in S-framed languages. This is
exemplified in (5) and (6) below with examples from Swedish and French, two
languages cited as candidates as S- and V-framed, respectively.
(5) En kvinna gå-r in i grotta-n.
DET.INDF woman walk-PRS in i cave-DET.DEF
‘A woman is walking into a cave.’
(6) Une femme entre dans la caverne.
DET.INDF woman enter.3SG.PRS in DET.DEF.Fcave
‘A woman exits the cave.’
Subsequent studies of the expression of motion have raised several objections to
dividing languages in two discrete types (e.g., Levinson and Wilkins 2006;
Bohnemeyer et al. 2007). One objection concerns languages where constituents
with equal grammatical and syntactical status express Manner- and Path-infor-
mation. Thai typically expresses Manner and Path as well as Deixis in three
separate verbs joined together in a serial-verb construction (Zlatev and
Yangklang 2004; Slobin 2004), as in (7).
(7) Phûyǐng doen khâw pai thâm.
woman walk enter go cave
‘A woman walks this way into a cave.’
(lit. ‘a woman walks exits goes cave’)
The questions of how many language types there are, or if there are construction
types rather than language types has been much debated (e.g., Croft et al. 2010;
The expression of non-actual motion 659
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
Beavers et al. 2010). However, independently of this question, the expression
of motion has been found to differ across languages, making it possible to
distinguish between at least three different types (Slobin 2004). I will therefore
refrain from entering the debates over construction type/language type, and
rather investigate to which extent the expression of physical (or actual) motion
affects the possibility to use motion verbs for describing static extensions in
space.
The paper is structured as follows. The concept of non-actual motion is
described in more detail in Section 2, together with Blomberg and Zlatev’s
re-interpretation of fictive motion (Talmy 2000a), subjective motion (Langacker
1990, 2006) and mental simulation of motion (Matlock 2004a, 2004b, 2010).
I describe the empirical study in Section 3 together with the semantic framework
of Holistic Spatial Semantics (Zlatev 1997). This framework was used to analyze
non-actual motion in Section 4 and the quantitative comparisons in Section 5. To
anticipate the results, speakers of all three languages exhibited a strong ten-
dency to use NAM-expressions for describing static spatial extensions. While
this result supports strong experiential and cognitive motivations behind this
phenomenon, speakers did so with resources closely related to the expression of
physical (or actual) motion. This suggests that pre-linguistic motivations are also
constrained by language-specific conventions. I conclude in Section 6 with a
summary and a tentative typology for the expression of non-actual motion. I
claim that this typology can be used to classify languages in an implicational
hierarchy of non-actual motion.
2 Theoretical background
Several different terms in the literature denote the use of motion verbs for
describing static situations. Langacker (1990) and Matsumoto (1996) speak of
subjective motion, Talmy (1983) of virtual, later fictive motion (Talmy 2000a),
while Matlock (2010) has used the term abstract motion. Despite their agreement
on motion being mentally superimposed motion on a static situation, the terms
fictive motion, subjective motion and abstract motion have different ranges of
meaning and justifications. The sentences in (8)–(11) are all analyzed as
instances of fictive motion by Talmy, based on “a cognitive bias towards
dynamism”(Talmy 2000a: 101). In contrast, Langacker (1990) restricts the
notion of subjective motion to the sentences in (8) and (10), which share “the
movement of the speaker’s focus of attention along a visual mental trajectory”
(Blomberg and Zlatev 2014: 389). Matlock (2010) characterizes (8)–(10) as
660 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
examples of abstract motion connected to mental simulation as a general
cognitive mechanism.
1
(8) The highway crawls through the city. (Matlock 2004a: 232)
(9) The milk is about to go sour. (Langacker 1990: 155)
(10) An ugly scar extends from his elbow to his wrist. (Langacker 2001: 9)
(11) The enemy can see us from where they are positioned. (Talmy 2000a: 115)
As can be seen, there is a terminological confusion where the scope of
previous analyses varies between different authors. The varying terminology
can be seen as an indication of the non-unitary nature of the phenomenon as
such, which in part might be a reason why it has proven difficult to semanti-
cally delimit as well as pinpointing a single explanation. The hybrid and
multi-motivated character of NAM has previously been noted by Matsumoto
(1996: 137), but without further exploring the implications: “[i]n some cases it
is the movement of the focus of attention; in other cases the motion of some
imaginary entity is involved; and in still other cases the mover is a specific
person (e.g., a speaker or a hearer)”. This terminological plurality can be
constructively re-interpreted as pointing to several experientially distinct moti-
vations with some type of sentences more closely linked to one kind of
experience than other.
2.1 Three experiential motivations
In order to investigate the multi-faceted nature of NAM, the possible motivations
should be clearly spelled out and differentiated from one another. The first
motivation is closely related to the capacity of self-motion and the experience
of movement and motility it brings with it. The ability to move is an indispen-
sable part of our experience, without which it would not be possible to grasp a
pen just out of reach, or explore the backside of a building. Our motility is
indispensable to the degree that without it, we could expect the world to appear
in a radically different way.
1The term abstract motion is also used by Langacker (1999: 82–83), but then to refer to
sentences where a motion expression stands for non-spatial changes (e.g., [9]).
The expression of non-actual motion 661
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
For a being completely immovable there would be neither space nor geometry; in vain
would exterior objects be displaced about him, the variations which these displacements
would make in his impressions would not be attributed by this being by change of
position, but to simple changes of state; this being would have no means of distinguishing
these two sorts of changes, and this distinction, fundamental to us, would have no mean-
ing for him. (Piaget and Inhelder 1956: 248)
To move, one must perceive that it is possible to move. For this reason, the
experience of movement involves the detection of affordances for motion. The
term affordances was introduced by Gibson (1977, 1979) to describe the intimate
connection between perception and action. In perceiving for instance a door
handle, it is not only perceived as a protruded part of a door, but primarily
(given previous experiences with such an object) as something which allows for
the door to be opened and closed. In other words, a constitutive feature of
perception is related to detecting the potential actions that can be carried out
with the physical environment. Similarly, roads, paths and trails are created
artifacts imbued with the affordance for (human) translocation. Roads are
manufactured to afford motion; trails are made in walking and over time
signaling a safely traversable route. The possibility for self-motion as one
motivation for NAM-sentences is further supported by the comparison between
Japanese and English by Matsumoto (1996). Japanese exhibits constraints on the
motion verbs that can express NAM for objects without affordance for human
translocation, see (12) and (13).
(12) Sono haiuee wa heeya no mannaka o
the highway TOP plain GEN center ACC
{tooru/iku/tootte iku}
{go.through/go/go.through.go}
‘The highway {goes through/goes in/goes through} the center (or middle)
of the plain.’
(Matsumoto 1996: 214)
(13) Sono densen wa heeya no mannaka o
the wire TOP plain GEN center ACC
{tooru/*iku/??toote iku}
{go.through/go/go.through go}
‘The wire {goes through/goes in/goes through} the center (or middle) of the
plain.’
(Matsumoto 1996: 215)
662 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
A range of different Japanese motion verbs that are acceptable in situations such
as (12) are unacceptable in (13), but not vice versa. We can then agree that the
lexical restriction is “cognitively reasonable, given the difficulty in evoking a
sensorimotor experience of moving along such a path in the mind of the
conceptualizer”(Matsumoto 1996: 215).
The affordance for motion is, however, just one of several possible experi-
ential motivations. NAM-sentences in English are insensitive to affordance for
motion, as in (14) where the Figure affords motion and in (15) where it does not.
Whether English or Japanese is exceptional in this regard is very much a
question for empirical research.
(14) The road goes through the forest.
(15) The wire goes through the forest.
This fact is often emphasized in previous analyses: both Talmy (2000a) and
Langacker (1990) turn their attention specifically to objects without affordance
for motion. In his analysis of fictive motion, Talmy (2000a) goes as far as aiming
to leave out objects with reference to actual motion.
2
A purer demonstration of this type of fictive motion would exclude reference to an entity
that supports the actual motion of other objects or that itself may be associated with a
history of actual motion. (Talmy 2000a: 104)
In his account of mental scanning as the motivating factor, Langacker (1990)
proposes that there is a kind of dynamicity in the very act of experiencing. The
act of “building up to a full conception”(Langacker 2001: 9) is disposed towards
dynamicity. As we perceive a moving entity, it is directly connected to a
continuous shift in our attention. This involves moving our own bodies, tilting
our heads, saccadic eye movements, and so forth. Dynamicity does not only
belong to the moving entity, but also to the act of perceiving. In (16), the Figure
is moving, but correlated with the very fact of grasping this situation is the
dynamism and movements of the observer. According to Langacker (1990,
2006), this dynamic aspect of conceptualizing is retained in a sentence such
as (17) to describe situations without any externally moving object.
2Talmy’s analysis of fictive motion is exceptionally detailed and involves a number of different
subtypes. If not otherwise noted, I am limiting my account to what Talmy refers to as coexten-
sion path, defined as “a depiction of the form, orientation, or location of a spatially extended
object in terms of a path over the object’s extent”(Talmy 2000a: 138).
The expression of non-actual motion 663
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(16) The balloon rises. (Langacker 2006: 25)
(17) The trail rises steeply near the summit. (Langacker 2006: 25)
Langacker (2001: 10) thus proposes that there is an inherent temporality and
dynamicity in experiential acts. It takes, so to speak, time to experience, to
conceive and to conceptualize. One indication in support of this view would be
using motion verbs for describing static situations. From this argument, we can
extract a second motivation for NAM-sentences, viz. the configuration of an
entity to allow for scanning. The predisposition to use NAM-sentences is
according to this motivation, insensitive to whether an extended object is
travelable or not.
Matlock (2004a, 2004b, 2010) argues that motion verbs for describing static
situations can be explained in terms of mental simulation. Most generally, mental
simulation proposes that “mental processes are supported by the same processes
that are used for physical interaction, that is, for perception and action”(Pecher
and Zwaan 2005: 1). Applied to language, this has lead to theories such as
simulation semantics, which proposes “that the meanings of words and of their
grammatical configurations are precisely the contributions those linguistic ele-
ments make to the construction of mental simulations”(Bergen 2007: 278). One
indication of mental simulation is that motion is intrinsically involved in under-
standing or producing expressions where a static extension is described with a
motion verb. On the basis of psychological experimentation, Richardson and
Matlock (2007) reach the following conclusion:
[I]n understanding an FM [fictive motion]-sentence, people re-activate and simulate
aspects of the protagonist’s motion, including speed, distance, and the terrain across
which the movement occurred. In doing so, they construct a dynamic representation that
mirrors the actual motion of the protagonist. (Richardson and Matlock 2007: 238)
As indicated by this quotation, simulation of motion involves re-activating
quite specific facets of a “protagonist’sactualmotion”. By virtue of involving
specific information about actual motion, the appeal to mental simulation can
be seen as complementing the accounts of affordance and visual scanning.
Self-motion and scanning is associated with actual motion: the possibility of
moving along a spatial entity and the dynamism of following a static object in
visual perception, respectively. In contrast to these two motivations, simula-
tion of motion (or perhaps more accurately, the imagination of motion)
requires the speaker to construe information about actual motion. Such a
motivation could be involved in NAM-sentences with motion verbs rich with
664 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
specific Manner-information involving, for instance, different postures, speeds,
bodily movements and so on.
3
In short, verbs with lexical semantics that refer
to movements carried out by someone or something. Examples of such uses are
showninexamples(18)–(20).
(18) The highway crawls through the city. (Matlock 2004b: 232)
(19) Insanity runs in my family…It practically gallops! (Arsenic and Old Lace,
cited in Brandt 2009)
(20) The dark velvet ditch creeps by my side. (T. Tranströmer, April and
Silence)
The ditch in (20) is not described as just moving by creeping, but rather resorts
to the “as-if”structure of imagination. The sentence does not point to the ditch
as affording self-motion, or to its extension in visual perception; rather it evokes
associations with movement patterns characteristic of certain animate beings
and particular bodily positions. Conceptualizing its meaning thus requires the
speaker or listener to engage in the activity of visualizing motion, which makes
this a distinct motivation.
4
In sum, three different and experientially distinct motivations to NAM-
sentences can be distinguished. The first of these is connected to self-motion,
and the affordance for motion. Per this view, the ability of the self to move along
entities affording motion would motivate the linguistic expression of NAM. The
second motivation is the dynamic character of (visual) experience and concep-
tualization, which can be seen as intimately connected to motion. Third, the
imagination of motion motivates NAM-sentences, especially for sentences richer
in Manner-information. At least these three experientially motivations can be
extracted from the literature, which in turn point to the multifaceted nature of
the phenomenon. By separating out the different motivations, it becomes possi-
ble to conduct empirical research on their respective role in the formation of
NAM-expressions across languages.
3Manner has been a notoriously difficult category to define coherently. Recent literature have
suggested distinctions between different kinds of Manner (e.g., Slobin et al. 2014). In this
article, I will not enter into a detailed discussion of how to define Manner.
4It is of course possible to have a vivid visualization of motion even when describing a non-
travelable entity with a more generic motion verb (such as go). However, this is not entailed
from the lexical semantics of the verb.
The expression of non-actual motion 665
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
3 Method
The present experiment was designed based on an operationalization of the
motivations discussed in Section 2. The first parameter was concerned with
spatially extended objects that afford motion and those that do not. Thus, the
objects depicted in the stimuli either afforded human movement or not. The
second parameter was concerned with a difference between the notions of
“cognitive bias towards dynamism”and “visual scanning”. This difference in
experiential engagement was operationalized by visually representing the same
spatial situation with perspectival alterations: either the spatially extended
object was represented from a first-person perspective (1pp) or from a third-
person perspective (3pp). The 1pp-stimuli were assumed to provide a sense of
being-there, thereby possibly enhancing the degree of involvement (or rather,
minimizing the indispensable distance between expression and content in every
visual representation). 3pp-stimuli provided a distanced view from an anon-
ymous perspective. Taking the parameters of Perspective and Affordance for
motion together gave a two-by-two design, summarized in (i) and (ii) with the
four different stimuli types shown in Figure 1.
i. A first-person perspective inviting the observer to move along the spatially
extended object vs. a third-person perspective providing an opportunity to
scan the figure.
ii. Entities that support human motion vs. those that do not (e.g., bridges vs.
fences).
3.1 Material
The pictorial material was designed to depict linear, extended objects placed in
an ecologically likely spatial environment. The stimuli comprised 38 pictures (24
target pictures, 12 controls and 2 practice pictures). Both target and control
pictures represented static situations without animate agents. The target pictures
fitted the criterion of depicting linear, spatially extended objects related to a
reference entity. These were further designed to capture the operationalized
motivations described above. The pair-wise designed pictures depicted the
same objects but from different perspectives. The target pictures are described
in Table 1. The spatially extended object was always positioned against a
reference entity (e.g., a house). To ensure that the relative relation between
object and the reference entity would not bias the results, the location of the
reference entity differed systematically across the stimulus set. For 3pp-pictures,
it was placed to the left/right of the spatially extended object. In 1pp-pictures,
666 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
the reference entity was proximal or distal with respect to the viewer’s perspec-
tive. The control pictures were designed to represent static situations without, or
with much less, linear extension than the target objects, such as park benches
and trees.
3.2 Participants
16 speakers of Swedish (8 female, mean age 28.4, sd = 9.5), 14 speakers of Thai
(10 female, mean age 29.9, sd = 7.9) and 13 speakers of French (11 female, mean
age 25.0, sd = 5.3) participated in the study. The Swedish and Thai participants
were recruited primarily through Lund University and compensated with a
voucher for their participation. The Thai participants were mainly exchange
students with less than two years spent in Sweden. French participants were
Figure 1: Stimuli according to the two parameters of Affordance and Perspective: (a) Afford/3pp.
(b) Afford/1pp. (c) Non-afford/3pp, and (d) Non-afford/1pp.
The expression of non-actual motion 667
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
recruited at the Montrouge section of the Ecole Normale Supérieure, Paris. In all
cases, a researcher fluent in the participants’native language conducted the
elicitation.
3.3 Procedure
All participants viewed the same 38 pictures in three different viewing orders.
Each order was viewed by as close to a third of the participants as possible. The
stimuli were presented on a laptop with a 15.4′’ widescreen monitor. The
researcher provided instructions in the native tongue of the participant and
controlled the presentation of stimuli through an external keyboard. The session
began with instructions in the participant’s native tongue, freely translated into
English as follows.
Table 1: Description of the target pictures.
Third-person perspective First-person perspective
Affords motion Affords motion
Object Reference
entity
Location (of
reference
entity)
Object Reference
entity
Location (of
reference entity)
Trail House Left end Trail House Trail begins at house
Trail House Right end Trail House Trail ends by house
Road Tunnel Left end Road Tunnel Road extends
outside tunnel
Road Tunnel Right end Road Tunnel Road extends inside
tunnel
Bridge Canyon Under bridge Bridge Canyon Under bridge
Row of stones River Surrounds Fig Row of
stones
River Surrounds Figure
Does not afford motion Does not afford motion
Fence Tree Left end Fence Tree Fence extends from
house
Fence Beach Right end Fence Beach Fence extends to sea
Pipe Tunnel Left end Pipe Tunnel Pipe extends
outside tunnel
Pipe Tunnel Right end Pipe Tunnel Pipe extends inside
tunnel
Phone wire Logs Under wire Phone wire Logs Under wires
Row of chairs Beach Surrounds Fig Row of chairs Beach Surrounds Figure
668 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
You will see a number of pictures. Look at each picture and describe what you see in one
sentence. Please try to provide natural and colloquial descriptions, as if you were to briefly
describe the picture to someone unacquainted with it. Some pictures are similar to each
other. Please describe each picture without referring to previously seen pictures. The
session begins with two warm-up pictures.
The instruction “describe what you see in one sentence”were used to ensure
that participants provided descriptions comparable in length and content. The
study was conducted in a silent room. Participants first viewed and described
two warm-up pictures, of which one represented actual motion: a car driving on
a road. A pre-test debriefing followed in which the researcher ensured that the
task was clear to the participant. The remainder of the session was recorded for
audio and video. Upon completing the elicitation task, a structured post-test
interview followed. Several participants expressed difficulty in describing the
pictures in just one sentence. They did not report this as a difficulty in following
the instructions, but found a single-sentence description to be insufficiently
detailed to exhaustively describe the picture. No participant reported the actual
purpose of the study.
3.4 Analysis
Recordings were transcribed exhaustively (except for noises, interruptions and
comments) in the standard orthography of the respective language. All descrip-
tions were segmented into clauses and all word-forms were provided with a
lexeme, an English gloss and a morpho-syntactic code.
The analysis required an operational definition of NAM-sentence. In order to
let the connection to actual motion remain explicit, a NAM-sentence was given
the following operationalized definition: a sentence that by substituting the
expression of the spatially extended object (typically expressed by the grammatical
subject) with one that denotes an entity that can move by its own volition, actual
motion would be the only possible interpretation. This is shown in (21) and (22),
with the (a) NAM-sentences corresponding to (b) actual motion-sentences by
substituting only the Figure-expressing nominal.
5
5An anonymous reviewer points out that the sentences in (22b), (23b) and (24b) would in
English be more idiomatic in the present progressive tense (i.e., is going). By virtue of signaling
that actual motion can be an ongoing activity and non-actual motion not, this could be an
additional marker of the difference between actual and non-actual motion. While the proposed
definition of NAM-sentences still holds for the three languages investigated, a refined definition
is a prerequisite for further research.
The expression of non-actual motion 669
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(21) a. A road goes into a tunnel.
b. Mary goes into a tunnel.
(22) a. The fence goes along a meadow.
b. John goes along a meadow.
Given this criterion, sentences with verbs expressing dynamism and change,
e.g., begin,continue and end were also included. From the point of view of
lexical semantics, these verbs do not only or primarily express Motion, but can
rather be analyzed in terms of continuity and dynamicity intimately connected
to motion (see Langacker 1990), or what Talmy (2000a: 257) calls “windowing of
attention”. To motivate their inclusion, the same test is applied in (23).
(23) a. The pipe continues out of the tunnel.
b. Jane continues out of the tunnel.
Several verbs also contain reference to dynamicity. Implied in verbs describing
posture or configuration is that the described state has come about by a prior
motion of the Figure. A verb such as delimit points to the fact that the Figure is in
a certain state, as in (24a). When used with an animate agent the sentence conveys
the activity of attaining this state, as in (24b). Still –this is not a description of
motion leading to a change of location –unlike the examples in (21b), (22b) and
(23b). For this reason, such sentences were excluded from the analysis.
(24) a. The fence delimits the beach.
b. Mark delimits the beach (by putting up a fence).
All descriptions were coded according to the framework of Holistic Spatial
Semantics (HSS). This theory proposes seven universal semantic categories for
spatial meaning (Zlatev 2003; Blomberg 2014). To these seven, Manner was
added as an eight category required to sufficiently capture the expression of
motion (actual and non-actual) across languages.
–Figure: The entity whose spatial location is profiled in the expression. Also
called Trajector (Langacker 1987) and referent (Levelt 1996).
–Landmark: A reference entity against which the (trans)location of the Figure
is gauged. Also called Ground (Talmy 2000b), and relatum (Levelt 1996).
–Frame of Reference: A stable and fixed perspective of space presupposed
by all spatial utterances (see Levinson 2003; Zlatev 2007; Blomberg 2014).
–Motion: The presence or absence of perceived motion. Can either express
change of location (e.g., enter) or the process of movement (e.g., jump).
670 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
–Region: A specification of a particular region of space with respect to the
Landmark, such as inside/outside the Landmark. This is similar to the
notion of search domain (Langacker 1987).
–Path: Defined schematically as the Beginning/Middle/End in an expression of
change-of-location, e.g., prepositions such as from/through/to, respectively.
6
–Direction: In contrast with Path, an expression of unbounded spatial
change-of-location, e.g., rise/fall,come/go and to the East.
–Manner: Involving different specifications of how the Figure moves, e.g.,
walk,run,jump,glide, hurry,spin, etc. This information can be related to
Path/Direction (i.e., change of location) or not.
Linguistic forms realize the semantic categories in different ways. While one form
often expresses one category, there are two other patterns. First, several semantic
categories can be expressed in a single form. For instance enter conflates Motion
with Path. Two or more distinct forms can express a single category (distribu-
tion), for example out of where Path is expressed in both an adverb/particle and
in a preposition. Due to these patterns, HSS proposes that meaning must be
analyzed on an utterance level. By taking the interplay between parts and wholes
into consideration, the meaning of a form depends partly on the linguistic
context in which it participates, and vice versa. Of specific relevance for the
analysis of NAM-sentences is that the meaning of motion verbs depends to a
certain extent on the surrounding linguistic context. As seen in examples (21)–
(24), whether the Figure can move or not participates in marking the difference
between actual and non-actual motion. This holistic perspective on meaning has
guided the semantic analysis presented in the following section. Furthermore,
these semantic categories are mapped to sentential constituents differently across
languages (Zlatev 1997, 2003). While the semantic resources are universal, they
are realized according to language-specific conventions.
4 The semantics of non-actual motion
4.1 Non-actual motion in Swedish
Motion-verb types were counted in the data and their semantic category was
determined on the basis of what type of actual motion the verbs expressed.
6The difference between Path and Region is similar to the distinction between conformation
and vector in Talmy (2000b).
The expression of non-actual motion 671
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
Results showed that the Swedish participants used 11 different Motion-verbs
distributed over five different semantic categories, as shown in Table 2.
Considering the semantic content of these verbs, they were either generic,
such as gå ‘go’and leda ‘lead’, or did not primarily serve as verbs for Motion,
such as sluta ‘end’and försvinna ‘disappear’. The semantically general verb gå
‘go’can express change in time, possession and extend to virtually any domain.
However, the basic meaning is a walking gait (i.e., as the English verb walk). It is
then notable that its use in (25) cannot be substituted for similar Manner-verb,
see (27). Limited to spatial extensions, the meaning of gå seems to be closely
related to another Swedish verb: leda ‘lead’. The latter verb can be used to
convey actual motion with meanings such as ‘to make something move’,‘to
steer’and ‘to direct’(for instance steering an animal, such as a horse or goat).
Outside of Motion, it can be used to express the consequence of a certain
behavior, process or intention (28) or to be in charge (29).
(25) En väg som gå-r in i en tunnel.
DET.INDF road COMP.REL go-PRS in in DET.INDF tunnel
‘A road that goes into a tunnel.’
(26) Avgränsning på strand-en som led-er ner till havet
delimitation on beach-DET COMP.REL lead-PRS down to sea-DET.DEF
‘Delimitation on the beach that leads down to the sea.’
(27) ? En väg som {vandra-r/promenera-r/}ini
DET.INDF road COMP.REL {wander-PRS/stroll-PRS/} in in
en tunnel.
DET.INDF tunnel
‘A road that {wanders/strolls/saunters} into a tunnel.’
Table 2: The different verbs used by the Swedish participants.
Manner
Path
Direction
Cause
Other
Total
flyta sluta komma leda sticka
‘float’‘end’‘come’‘lead’‘stretch’
gå försvinna binda fortsätta
‘go’‘disappear’‘bind’‘continue’
löpa
‘run’
spruta
‘spurt’
672 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(28) Snabb-a lösninga-r kan leda till katastrof.
Quick-PL solution-PL can lead.INF to catastrophe
‘Quick solutions might have disastrous effects.’
(29) General Custer leder en armé.
General Custer lead-PRS DET.INDF army
‘General Custer commands an army.’
The Swedish participants also used the Manner-verb löpa ‘run’, as in (30). As a
Motion-verb, it conveys a sense close to springa ‘run’.
(30) Ett trä-staket som löper i rät linje.
DET.INDF wooden-fence COMP.REL run.PRS in straight line
‘A wooden fence that runs in a straight line.’
This seems to suggest that Swedish NAM-sentences can retain information about
the Manner of motion. However, Manner- and Cause-verbs can be used as long
as this information is demoted. Apart from the pre-given intuition that informa-
tion about pace, gait, etc. should be demoted in NAM-sentences (Matsumoto
1996), it is not really clear wherein the demotion of Manner-information lies.
If we extend the scope to entire clauses, it is possible to formulate this intuition
more clearly. In (25), (26) and (30), the entities are described together with
adverbs and prepositions contributing to the expression of Motion. In other
words, the demotion lies in that the Path- and Direction-information given
about a static object “overrides”the Manner-information provided by the
verb.
7
This is one explanation why it typically sounds strange to qualify
Manner in a NAM-expression: if the Manner-verb in (30) is modified with the
adverb snabbt ‘quickly’, then the sentence is semantically deviant, as shown in
(31). Alternatively, the verb löper might be associated with the mental scanning
of the wooden fence, rather than the object itself.
8
Given such an explanation, it
would be equally strange to conceive of the mental scanning as something that
can be qualified in terms of adverbs such as ‘quickly’.
7A different way to phrase the analysis of Path/Direction “overriding”Manner would be in
terms of basic meaning changing to contextual meaning. That is, a verb that expresses Manner
as its basic meaning can have contextual meanings where Manner is not expressed. One
example of this would be NAM-sentences. I owe this important remark to one of the anonymous
reviewers.
8I would like to thank an anonymous reviewer for pointing out this alternative explanation in
terms of mental scanning.
The expression of non-actual motion 673
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(31) ? Ett trä-staket som löp-er snabbt i rät linje.
DET.INDF wooden-fence COMP.REL run-PRS quickly in straight line
‘A wooden fence that runs quickly in a straight line.’
If we understand NAM-sentences as demoting Manner-information across the
clause, then it is possible to pinpoint why some sentences might be in need of
an imagination-based interpretation. As discussed in Section 2, neither affor-
dances nor visual scanning can account for NAM-sentences rich on Manner-
information. With the addition of more Manner-information, either with an
adverb or with verbs expressing more elaborate forms of Manner, comprehen-
sion cannot resort to conventional semantics. We could even generalize to the
point of saying that the possibility to participate in sentences where Manner-
information is overridden is essential for a Motion-verb in order to be
regarded as “bleached”. The type of sentences and constructions that partici-
pate in expressing NAM can be expected to typically include verbs and
constructions already applicable to a wide array of domains. Other examples
attested in the data include fortsätta ‘continue’and sluta ‘end’, i.e., verbs that
by highlighting an aspectual phase can express the continuation or end of a
process in general.
The Swedish participants also provided an additional interesting type of
descriptions that by definition fall outside of NAM-sentences. These involved
sentences without motion verb –in fact with no verb at all –but with the
adverbs ut and ner (‘out’and ‘down’, respectively), as shown in examples
(32)–(34). In these complex NPs, the adverb was combined with a preposition.
(32) En landsväg ut genom en tunnel.
DET.INDF road out through DET.INDF tunnel
‘A road out through a tunnel.’
(33) Ett rör ut genom en tunnel.
DET.INDF pipe out through DET.INDF tunnel
‘A pipe out through a tunnel.’
(34) Ett staket på en strand ner mot vattnet.
DET.INDF fence on DET.INDF beach down towards water.DET.DEF
‘A fence on a beach down towards the water.’
It is not possible to express actual motion in this way, see (35). What could be
expressed, however, is something like the view from a window, as in (36), where
674 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
the window is presented as facing towards the sea.
9
Arguably, however, what is
directed toward the sea is the vantage point of a (possible) perceiver. Standing
by the window, one is provided with a framed perception of the sea. In a
building, windows are for looking out; in a sense they are there to afford and
to allow a view of the outside. Thus, to look in through a window is to be a
Peeping Tom, unless, of course, it is a display window.
10
(35) * En man ut mot strand-en.
DET.INDF man out towards beach-DET.DEF
‘A man out towards the beach.’
(36) Et fönster ut mot hav-et.
DET.INDF window out towards sea-DET.DEF
‘A window facing towards the sea.’
In (36), the situation is different from examples (32)–(34). The window in (36), in
contrast to a highway or a pipe, does not occupy the space between the house
and the sea. What remains to be investigated are the cross-linguistic tendencies
to express perceptual vantage points in the same way as spatial extensions. If
found to be a stable pattern, this could be read as support for an extended
reading of Langacker’s notion of visual scanning: perception, even in the
absence of a scan-able object, searches for a correlate that can stand in for the
act, such as a window.
In sum, all picture types were found to elicit NAM-sentences in Swedish with
generic or bleached verbs as gå ‘go’and leda ‘lead’together with adverbs and
prepositions. Using the resources for expressing actual motion, the semantic
focus of the Swedish descriptions was how the Figure as a spatially extended
object was related to the spatial environment. This sense was conveyed in two
additional ways:
i. Change-of-state verbs not restricted to motion, e.g., börja ‘begin’and sluta
‘end’+ Path-and Direction prepositions and/or adverbs.
ii. Complex NPs with only prepositions and adverbs expressing Path and/or
Direction.
9As Talmy correctly notes, (36) differs from the previous sentences. The path is not formed
through a physical objects, but from the gaze of visual perception, what Talmy (2000a) calls
prospect path.
10 My thanks to Göran Sonesson for pointing out the cultural affordances of windows.
The expression of non-actual motion 675
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
Per the definition of NAM-sentences as “a sentence which in principle could
describe actual motion, and by substituting the Figure-expression with one that
denotes an entity that can move by its own volition”, the sentences falling under
(i) are instances of NAM-expressions. In contrast, (ii) cannot be used to convey
the sense of actual motion and is therefore best considered a semantic sibling to
Non-actual motion, which I call Non-actual Path.
4.2 Non-actual motion in French
The descriptions produced by French speakers were similar to the Swedish ones
in that the French speakers also tended to use bleached and generic Motion-
verbs for expressing NAM. The range of used motion verbs types was, however,
wider, including 35 different verbs (see Table 3).
The generic motion verbs mener ‘lead’and aller ‘go’realized NAM-sentences
across all stimuli types, as shown in (37)–(39).
11
Table 3: The different verbs used by the French participants.
Manner
Path
Direction
Cause
Manner+Path
Other
Total
baigner entrer diriger conduire jaillir aller
‘bathe’‘enter’‘direct’‘lead’‘spurt out’‘go’
écouler passer longer mener s’enfoncer bouger
‘flow’‘pass’‘run along’‘lead’‘penetrate’‘move’
fuir sortir monter relier pénétrer deboucher
‘flee’‘exit’‘ascend’‘connect’‘penetrate’‘open into’
marcher traverser venir sèparer se poursuivre
‘walk’‘cross’‘come’‘separate’‘continue’
s’étendre arriver se diriger transporter
‘expand’‘arrive’‘head for’‘carry’
retourner suivre emprunter
return’‘follow’‘take’
partir s’avancer prendre
‘leave’‘move ‘take’
accéder forward’continuer
‘reach’s’approcher ‘continue’
‘go closer’
11 The glossing of the French examples clearly deviates from standard French orthography, for
instance by separating forms that are typically contracted both in speech and writing, such a
d’un -> de un.
676 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(37) Un petit chemin qui mène à la porte.
DET.INDF.Msmall path COMP.REL lead.3SG.PRS to DET.DEF.Fdoor
‘A small path that leads to the door.’
(38) Une barricade qui va jusqu’àle horizon.
DET.INDF.Fbarricade COMP.REL go.3SG.PRS until DET.DEF.Mhorizon
‘A fence that goes all the way to the horizon.’
(39) Une barrière qui mène jusqu’àla mer.
DET.INDF.Ffence COMP.REL lead.3SG.PRS until DET.DEF.Fsea
‘A fence that leads all the way to the sea.’
We can apply the same analysis to these NAM-sentences as to the Swedish ones:
the verb together with a preposition expresses the continuity of the spatial
extension.
As could be expected, Path-verbs were recurrent in the French data. This
can be seen as clearly related to the typological characterization of French as a
verb-framed language. Shown in (40)–(41) are NAM-sentences that follow the
common expression of motion in French (see Blomberg 2014).
(40) Une route qui passe sous un tunnel.
DET.INDF.Froad COMP.REL pass.3SG.PRS under DET.INDF.Mtunnel
‘A road that passes under a tunnel.’
(41) Une route qui traverse une montagne.
DET.INDF.Froad COMP.REL cross.3SG.PRS DET.INDF.Fmountain
‘A road that crosses a mountain.’
When the Figure was partially inside and partially outside the Landmark, the
French speakers tended to use the verbs sortir ‘exit’and entrer ‘enter’. These
verbs were used analogously to their use in actual motion descriptions, i.e.,
together with prepositions de ‘from’and dans ‘in’, respectively, see (42) and (43).
It is notable that in (42), the speaker uses the gerund form en sortant. This
progressive form suggests the situation is construed as-if the speaker rather than
the road is exiting the tunnel. Even though such constructions were scarce in
the data, they might be an additional indication of an imagination-based
motivation.
12
12 My thanks goes to an anonymous reviewer for making me aware of the progressive form as
possibly indicating an imagination-based explanation.
The expression of non-actual motion 677
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(42) Là c’est en sortant de un tunnel.
here 3SG be.3SG.PRS in exit.PRS.PTCP from DET.INDF.Mtunnel
‘Here we are leaving a tunnel.’
(43) Les canalization-s qui rentre dans un mur.
DET.DEF.PL pipe-PL COMP.REL enter.3SG.PRS in DET.INDF.Mwall
‘The pipes that enter in a wall.’
Apart from these expected patterns of Path-verbs and other Motion-verbs applic-
able to a wide array of domains, the French speakers exhibited some additional
patterns. One of these concerns sentences that express Direction rather than
Path, as in (44) and (45).
(44) Une route qui se dirige vers
DET.INDF.Froad COMP.REL PRON.REFL head.for.3SG.PRS toward
une forêt.
DET.INDF.Fforest
‘A road that heads toward a forest.’
(45) Un chemin de pierre en plain milieu de l’eau
DET.INDF.Mpath of stones in.the.middle. of DET.DEF.Fwater
qui longe une rivière.
COMP.REL run.along.3SG.PRS DET.INDF.Friver
‘A path of stones in the middle of the water that runs along a river.’
There were also instances of verbs conflating Manner and Path, e.g., pénétrer
‘penetrate’. This can be interpreted as expressing the process through which the
entity came to be, in the present study restricted to pipes and roads through
tunnels (46). To attain the state-of-affairs, a previous force actually penetrating
or drilling through the landscape is required. The reference to the prior process
is retained in describing the present landscape.
(46) Cette belle conduite qui semble pénétrer
DET.DEM.Fpretty pipe COMP.REL seem.3SG.PRS penetrate.INF
sous ces roche-s.
under DET.DEM.PL rock-PL
‘This pretty pipe that seems to penetrate under those rocks.’
Arguably, the function of an object can make the differentiation between actual
and non-actual motion somewhat blurred. For instance, a pipe can serve as a
678 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
conduit for transporting liquids. Even though the entity itself is static, it contains
something moving. In this way, there is a possible ambiguity in the motivation
to this use of NAM-sentences: either the function of actually transporting water
or the spatial configuration of a pipe. Considering the multi-faceted nature of
NAM discussed in Section 2, it may be expected to be both.
4.3 Non-actual motion in Thai
In motion typology, Thai has been characterized as an equipollently-framed
language (Slobin 2004). This is due to the serial-verb constructions (SVCs) of
Thai with Path, Manner and Deixis expressible in separate verbs with equal
syntactic status (Zlatev and Yangklang 2004; Blomberg 2014).
13
Looking at the
type frequency of produced motion verbs, the Thai speakers were in-between the
Swedish and French participants (Table 4).
One of the first things to note is that the Thai speakers’NAM-descriptions
did not contain verbs corresponding to Swedish leda ‘lead’or French mener
‘lead’. While the verb phaa ‘lead’can be used to express actual motion, as in
(47), it cannot be used to express NAM.
(47) Phuying phaa dek pai rohngriian.
Woman lead child go school
‘The woman leads the child to school.’
Table 4: The different verbs used by the Thai participants.
Manner Path Direction Cause Manner+Path Other Total
doen khâw khuên chueam chò thai
‘walk’‘enter’‘ascend’‘connect’‘pierce’‘take’
phûng khaâm long riang ‘put lôt toò
‘dash’‘cross’‘descend’in order’‘penetrate’‘continue’
wîng oòk ma tàt thòt
‘run’‘exit’‘come’‘cut-through’‘take-off’
phaàn pai thálú
‘pass’‘go’‘go-through’
thu˘eng taam
‘reach’‘follow’
13 In the semantic analysis, deictic motion was analyzed as form of Direction. Specifically,
deictic motion was considered as expressing Direction according to a Viewpoint-centered Frame
of Reference.
The expression of non-actual motion 679
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
Instead, the Thai participants typically produced serial-verbs construction with
some constraints dependent on the type of entity described. The SVCs often
contained one Path- and one Deictic verb. In contrast to sentences describing
actual motion, NAM-sentences characteristically did not contain a Manner-verb
in the SVC (Blomberg 2014). Before venturing further into SVCs, let us begin
from sentences containing a single motion verb.
The Thai speakers used Path-verbs such as phaàn ‘pass’and khaâm ‘cross’.
Both these verbs were used to describe a bridge between two cliffs, see (48) and
(49). The first of these, phaàn, has a wider range in the data and can also be
used for pictures where the Figure is such that it typically passes through the
Landmark, such as a road or pipe passing through a tunnel, as in (50).
(48) Mi saphan yao phaàn maênaám.
COP bridge long pass river
‘A long bridge passes a river.’
(49) Mi saphan khaâm heǒ.
COP bridge cross ravine
‘A bridge crosses a ravine.’
(50) Thànǒnphaàn umong yù bon laì-khǎo.
road pass cave exist top hillside
‘A road passes a cave (that) is on the top of a hillside.’
As shown in (51)–(53), the Path/Manner-conflating verbs lôt ‘penetrate’,tàt ‘cut-
through’and chò ‘pierce’were attested in the data, the latter only in a serial-verb
construction with the two additional Path-verbs thálú ‘go-through’and phaàn
‘pass’. Just as in French, these Path/Manner-conflating verbs were only used to
describe roads and pipes entering and exiting tunnels. The first of these is
similar to the French verb pénétrer ‘penetrate’in that the verb could be used
to express the actual motion responsible for bringing about the state-of-affairs.
In order to make a tunnel, one drills through the mountain. Force is enacted on
the mountain; a force retained by describing the relation between tunnel and
mountain with verbs such as lôt and pénétrer.
(51) Pen thang lôt tai umong.
be way penetrate under cave
‘A road penetrates under a mountain.’
680 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
(52) Pen thàňon thî tàt phaàn phukhǎo.
be road COMP cut-through pass mountain
‘A road that cuts-through and passes a mountain.’
(53) Thang rotyon chò thálú phaàn phukhǎo.
way car pierce go-through pass mountain
‘A car road pierces through and passes a mountain.’
The relation between a Figure partially inside and partially outside a Landmark
could also be described with the verbs khâw ‘enter’and oòk ‘exit’,asseenin(54)
and (55).
(54) Thangdoen khâw pai nai umong.
path enter go inside cave
‘A path enters goes inside a cave.’
(55) Thàňon oòk ma chaàk umong.
road exit come from cave
‘A road comes out of a cave.’
Since both khâw ‘enter’and oòk ‘exit’mainly focus on the discrete transition
between two distinct Regions (Zlatev 2003), a sense of dynamic continuity can
be added not only by the deictic ma ‘come’and pai ‘go’,butalsobytheverbtoò
‘continue’.Similartofortsätta ‘continue’in Swedish, toò does not express Motion
per se, but in the context of NAM-sentences provides the continuity of the extended
object beyond the transition from inside to outside. This is shown in (57) and (58).
14
(56) Pen tho yao toò kan oòk ma chaàk umong.
be pipe long continue PRON.REFL exit come from cave
‘A long pipe continues out from a cave.’
(57) Mi tho thî toò kan yao khâw pai nai umong.
COP pipe COMP continue PRON.REFL long enter go inside cave
‘A pipe that connects together and enters a cave.’
The previous examples illustrate the use of deictic verbs together with Path-
verbs. However, both pai and ma occurred as the single verb in a NAM-sentence,
14 The phrase toò kan yao in (57) is a form of lexicalized phrase with a meaning not reducible
to its parts: ‘connects to make long’.
The expression of non-actual motion 681
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
see (58) and (60). Pai ‘go’also occurred together with other motion verbs, as
in (59).
(58) Mi thangdoen pai nai umong.
COP path go inside mountain
‘A path goes inside a mountain.’
(59) Mi tho sueng khuên pai.
COP pipe COMP ascend go
‘A pipe goes up.’
(60) Pen tho thî ma chaàk umong.
be pipe COMP come from cave
‘A pipe that comes from a cave.’
These sentences indicate that the uses of ma ‘come’and pai ‘go’differed, with
more constraints for ma ‘come’than for pai ‘go’. This difference was reflected in
the only type of stimuli that allowed ma were pictures from a first-person
perspective with the object extending outside the Refence entity (as shown in
Figure 2). A tentative explanation must first acknowledge the importance
of deictic information in Thai, which is where the motivation from experience
comes in. Confronted with the choice of pai ‘go’or ma ‘come’in a
Figure 2: Stimuli from First-person perspective where the Figure is extending beyond the
perceptual field of view.
682 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
NAM-situation, the former could be explicated in terms of Langacker’s (1990)
scanning analysis: when following a static object in visual perception, it could
be more preferably conceptualized as a non-actual movement departing from
the speaker rather than (non-actually) reaching the same. However, this does
not account for the use of ma ‘come’when describing stimuli such as in Figure 2.
Here, the Figure has no apparent continuity forward in space and the road
“disappears”into the darkness. It could be argued that the stimulus thereby
presented less motivation for pai ‘go’, implying departure from the speaker’s
perspective than for ma ‘come’, implying its arrival at the speaker’s location.
I have hitherto discussed Path and Direction (specifically Deixis) in NAM-
sentences. Given this characterization, it seems as if the Thai speakers behaved
somewhat similarly to the French participants, but with Deixis added. However,
such a conjecture does not take one of the most important the important features of
Thai motion expressions into account: SVCs and the recurrent pattern of combining
verbs expressing Manner, Path and Deixis. In NAM-sentences, this was reflected in
the form of three different Manner-verbs which occurred in SVCs: doen ‘walk’,wîng
‘run’and phûng ‘dash’, cf. (61)–(64). These were always found together with at least
one Path-verb, what Rojo and Valenzuela (2004) call path-related manner-verbs.
(61) Mi thang léklék doen khâw pai.
COP way small.small walk enter go
‘A small road goes (in)to [a house].’
(62) Thàňon sen nuèng sueng wîng khâw pai bon phukhǎo.
road CLF NUM COMP run enter go inside mountain
‘A long road that runs away into a mountain.’
(63) Pen thàňon thî wîng oòk ma chaàk umong.
be road COMP run exit come from cave
‘A road that runs out from a cave.’
(64) Mi thàňon phûng khâw su thang.
COP road dash enter to way
‘A road dashes into a way.’
The use of the Manner-verbs wing ‘run’,doen ‘walk’and phûng ‘dash’in Thai seems
to be highly motivated by the typical velocity of movement along the object. Paths
and small roads were described with doen ‘walk’, whereas larger roads such as
freeways, where the velocity is typically faster, were described with wîng ‘run’or
phûng ‘dash’. Manner-verbs thus convey information about how fast one tends to
The expression of non-actual motion 683
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
move on the object in question.
15
This suggests that Manner in Thai contributes
crucial information to NAM-sentences not seen in Swedish and French.
To sum up, the expression of NAM in Thai differed from Swedish and French in
several crucial respects. All three types of motion-verbs were used in expressing
NAM: Path and Direction-verbs independently or together, as well as Manner-verbs.
The Region-changing verbs khâw ‘enter’and oòk ‘exit’were used in contexts where
thefiguredidnot“enter”or “exit”a landmark. This suggests that the verbs are
bleached even in the context of NAM-sentences. Deictic verbs, specifically pai ‘go’
marked (at least) the vantage point of the speaker’s conceptualization. I have argued
that this use is both motivated from the experience of NAM and dependent on
linguistic conventions. The speakers’use of ma ‘come’was restricted to stimuli of
entities from a 1pp with spatially extensions partially located outside and partially
inside the reference entity (i.e., tunnel). This constraint was tentatively interpreted as
experientially motivated by the lack of a forward continuation of the entity in
question. In these pictures, the extended objects are not seen far beyond the tunnel
entrance, as shown in Figure 2. For this reason, there is nothing that is “away”from
the speaker as much as “towards”the speaker. Another tendency used only by the
Thai-speakers was Manner-verbs in NAM-sentences. Dependent on whether the
Figure was associated with fast or slow travel, the Thai speakers marked this
difference with Manner-verbs expressing different velocities of human translocation.
This contribution of Manner-verbs to NAM-sentences was not found in Swedish and
French, where Manner-verbs, if they occurred at all, were bleached and interchange-
able with Motion-verbs such as Swedish gå ‘go’or French aller ‘go’.
5 Quantitative comparisons
This section describes the quantitative results from the study, beginning from
overall proportions of NAM-sentences across the three language groups before
moving on to the distribution of verbs and verb types. Finally, the proportion of
NAM-sentences across the four experimental conditions is described and discussed.
5.1 Proportion of non-actual motion descriptions
First of all, it was important to determine if the descriptions in the three
languages were comparable. The proportions of descriptions that contained at
15 Of course, the Manner-information pertains to velocity and not the type of movement associated
with the entity in question. For instance, one typically drives and not runs,alongafreeway.
684 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
least one clause with any or several of the spatial semantic categories are shown
in Table 5.
As can be seen from this table, the speakers of all three languages produced
a high number of spatial descriptions for target as well as for control stimuli,
indicating that they performed according to the instructions. Approximately
40% of all spatial descriptions contained a NAM-description in all three lan-
guages, with the Thai and French groups producing slightly more on average
(see Figure 3). This could be compared with the control pictures, which as
expected, elicited NAM-descriptions in less than 5% for all three language
groups. As shown in a paired two-sample t-test, the difference between target
and control stimuli in the proportions of NAM-descriptions was statistically
significant (t = 15.0, df = 42, p< 0.001). The differences between languages were
Table 5: Percentages of descriptions with spatial information for target and control stimuli.
Swedish Thai French
Target .%(SD =.%) .%(SD =.%) .%(SD =.%)
Control .%(SD =.%) .%(SD =.%) .%(SD =.%)
ThaiFrenchSwedish
Control
Target
0.0
0.2
0.4
0.6
0.8
1.0
Figure 3: The proportion of NAM-descriptions in Swedish, French and Thai.
The expression of non-actual motion 685
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
not significant. This indicates that the participants clearly differentiated target
from control stimuli. Furthermore, the proportion of NAM-sentences is suffi-
ciently high to suggest that they are not only a possible strategy, but also
common and conventionalized in the three languages.
5.2 Distribution of verb types
With respect to the motion verbs used, speakers of all three languages displayed
a strong tendency to prefer general motion verbs with applicability to domains
other than motion and space (Table 6). The most common verbs were quite
sparse on information about Manner of motion. This is in line with observations
previously made by Matsumoto (1996) and Rojo and Valenzuela (2004): motion
verbs in NAM-sentences must retain information pertaining to Path of motion
while demoting information about Manner of motion.
Table 6: The most common Motion-verbs in Swedish, French and Thai.
Translation Occurrences Percentages
Swedish verbs
gå go .
leda lead .
komma come .
fortsätta continue .
French verbs
sortir exit .
mener lead .
traverser cross .
aller go .
rentrer (re-)enter .
passer pass .
Thai verbs
pai go .
khâw enter .
phaàn pass .
oòk exit .
ma come .
lôt penetrate .
tàt cut-through .
Percentages of total Motion-verb tokens in respective language.
686 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
The Swedish participants used the two verbs gå ‘go’and leda ‘lead’in almost 80%
of all cases. As noted in 4.1, these verbs are widely used in domains other than
actual motion. Still, that two verbs would make up almost 80% of all
NAM-sentences is quite noteworthy. In addition to these two verbs, the deictic
verb komma ‘come’and the fortsätta ‘continue’also occurred quite frequently. The
latter expresses continuity and signals aspectual phases (Talmy 2000a), but when
applied to NAM, it occurs together with Path/Direction adverbs and prepositions
and thus participates in expressing the spatial continuity of an entity’sextension.
Consistent with the larger amount of verb types, the French speakers were
also more diverse in terms of verb type frequencies. Different Path-verbs were
common, such as sortir ‘exit’and rentrer ‘re-enter’,traverser ‘cross’and passer
‘pass’. The two generic verbs mener ‘lead’and aller ‘go’also recurred.
16
The Thai
speakers regularly used the deictic motion-verb pai ‘go’–to a much greater extent
than its opposite ma ‘come’. Considering the dedicated slot for deictic verbs in
SVC, a high frequency of such verbs is to be expected. A partial explanation for
the higher frequency of pai ‘go’is that go-verbs are often less deictic than come-
verbs (see Fillmore 1997, and the discussion of visual scanning in Section 4.3).
This would suggest that pai is more semantically general than ma. The verbs khâw
‘enter’and oòk ‘exit’were recurrent in the data, as were phaàn ‘pass’,aswellas
the Path+ Manner-conflating verbs lôt ‘penetrate’and tàt ‘cut-through’.
As previously discussed, Swedish, French and Thai have been considered
as prototypical candidates for the three different types in motion typology,
satellite-, verb- and equipollently-framed languages (Slobin 2004; Zlatev and
Yangklang 2004). How, if at all, does this transfer to non-actual motion?
With very few exceptions, the experimental conditions Afford motion and
Perspective did not influence verb preference.
Among the few exceptions were the two most common verbs in the French
data, which exhibited differences dependent on the condition. The verb sortir
‘exit’occurred mainly when the depicted spatial extension did not afford human
motion, such as pipes (n = 20 for Non-afford, n = 6 for Afford). When the exten-
sion afforded human motion, sortir was elicited only for stimuli of figures drawn
from a first-person perspective. No similar constraints pertained to the opposite
direction, i.e., entrer ‘enter’. Instead of sortir, the French speakers preferred
verbs such as traverser ‘cross’and passer ‘pass’. This could possibly indicate
that continuity and travel is associated with roads but not with pipes. While
insensitive to Perspective, mener ‘lead’showed the opposite pattern with respect
to Affordance: 17 out of 19 occurrences were produced for Afford. The Swedish
16 Interestingly, Rojo and Valenzuela (2004) found Spanish speakers to use a more diverse set
of motion verbs than English speakers.
The expression of non-actual motion 687
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
verb leda ‘lead’exhibited a similar tendency accounting for 46 out of 62
occurrences for the condition Afford. While in part attributable to the higher
number of NAM-descriptions for this condition, the difference in occurrences is so
considerable that the higher number of NAM-descriptions can only constitute a
partial explanation. A possible explanation is that leda and mener are goal-directed:
they can express different forms of processes with a direction towards an end. It
could be possible that since roads afford human motion they are seen as imbued
with a telos. Fences and pipes, by contrast, are not made for human beings to travel
along. The production of the two most common verbs in Thai khâw ‘enter’and pai
‘go’combined in a serial-verb construction, khâw pai ‘enter go’, was also constrained
by the experimental conditions: 9 out of 11 occurrences were found in cases where
thefigureaffordedmotion.Theuseoftheseverbsastheonlyverbinaclause,or
together with other verbs in SVCs, was not restricted by conditions in the same way.
The general tendency to use a restricted set of motion verbs independent of the
experimental conditions speaks in favour of the conventionalized character of
NAM-expressions in the three languages. At the same time, the specific preferences
of some verb uses can be explained by appeal to the experiential motivations.
5.3 Non-actual motion sentences per experimental condition
We saw in Section 4 that NAM-descriptions were prolific, but we have yet to
determine how they were distributed across the four experimental conditions.
NAM-descriptions were most often used for the condition 1pp/Afford motion in
all three languages (Figure 4). This effect was strongest for the Thai speakers,
intermediate for the French group and weakest for the Swedish participants.
The data was analyzed with regression analysis with fixed and random
effects for two-way interaction between the experimental conditions. This ana-
lysis found that Afford Motion + 1pp differed significantly from the other condi-
tions (χ² = 11.8, df = 1; p< 0.0001). Three-way interaction with language as a third
factor was not significant (χ² = 13.2, df = 8; p= 0.10). The large differences
between control and target stimuli (see Figure 3) make a strong case for the
prevalence of NAM-sentences in the three languages. This could possibly be
explained by appeal to an explanation along the lines of visual scanning.
However, it cannot account for the significant differences between the experi-
mental conditions. The motivation from self-motion would explain why NAM-
descriptions were most strongly elicited under the conditions of Afford + 1pp.
Finally, the dominance of Afford + 1pp could also predicted by the imagination-
based hypothesis, but the actual verbs used (with the possible exception of the
Manner-verbs in Thai) do not support this notion.
688 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
In sum, the results support the multi-motivated character of NAM: all three
motivations were found to have a decisive role in the production of NAM-
sentences to different degrees. This proposes that NAM is a hybrid phenomenon
covering several experiential motivations conventionalized differently across
languages.
6 Discussion: Towards a typology of non-actual
motion in language
All stimuli types elicited NAM-descriptions in all three languages. From this, it
can be concluded that NAM-sentences are highly conventionalized in Swedish,
French and Thai. Speakers of all three languages described pictures in ways that
suggest a connection to motion, such as dynamism, aspectual phases, vantage
point, and so forth. Even if all three languages have the resources for expressing
NAM, they realize these differently in ways clearly reminiscent of how actual
motion is expressed in each respective language. The Swedish speakers used
(bleached or generic) Motion-verbs together with Path/Direction adverbs and
ThaiFrenchSwedish
Afford,3pp
Afford,1pp
Non– afford,3pp
Non– afford,1pp
0.0
0.2
0.4
0.6
0.8
1.0
Figure 4: Proportion of NAM-descriptions per stimuli type and language; error bars represent
standard deviation.
The expression of non-actual motion 689
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
prepositions. This naturally follows from the few Path-verbs in Swedish: among
the few examples are korsa ‘cross’and passera ‘pass’. These were not attested in
the present study. The French speakers produced both Path-verbs and generic
Motion-verbs and the Thai speakers produced serial verb-constructions Path-
and Deictic verbs and sometimes also with Manner-verbs. These variations
between languages indicate how the pre-linguistic experiences discussed in
Section 2 adapt to the internal logic of language-specific semantic conventions
and constraints.
Several different types of NAM-sentences were detected, some common to
all and some unique to one of the three languages. Both Swedish and French
participants used generic Motion-verbs to express NAM, as in (65) and (66).
These types of expressions were not sensitive to the stimuli described.
(65) En väg som gå-r in i en tunnel.
DET.INDF road COMP.REL go-PRS in in DET.INDF tunnel
’A road that goes into a tunnel.’
(66) Une barricade qui va jusqu’àl’horizon.
DET.INDF.Fbarricade COMP.REL go.3SG.PRS until DET.DEF.Mhorizon
‘A fence that goes to the horizon.’
Both the French and Thai speakers used Path-verbs, as in (67)–(68), including
enter/exit verbs; see (69)–(70).
(67) Mi saphan khaàm hěo.
COP bridge cross ravine
‘A bridge crosses a ravine.’
(68) Une route qui traverse une montagne.
DET.INDF.Froad COMP.REL cross.3SG.PRS DET.INDF.Fmountain
‘A road that crosses a mountain.’
(69) Thangdoen khâw pai nai umong.
path enter go inside cave
‘A path goes to the inside of a cave.’
(70) Les canalization-s qui rentrent dans un mur.
DET.DEF.PL pipe-s COMP.REL enter.3PL.PRS in DET.INDF.Mwall
‘The pipes that enter into the wall.’
690 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
These sentences are similar in that they use (a) generic Motion-verbs and/or
(b) the language-specific conventions for actual motion. From this baseline of
characteristic or typical NAM-sentences, there are related sentences where the
reference to motion is diminished, but which still (i) clearly use the vocabulary
and the constructions of motion and (ii) evoke the sense of motion. One clear
examples of this type were complex NPs with “dynamic”(Path/Direction) pre-
positions and adverbs in Swedish, as in (71).
(71) Ett rör ut genom en tunnel.
DET.INDF pipe out through DET.INDF tunnel
‘A pipe out through a tunnel.’
Since they evoke dynamism and change, sentences with a verb expressing
change of state or aspectual phases are semantically somewhat similar to the
complex NPs above. These verbs, which are found in all three languages, are
used in NAM-sentences to express the change of something immobile. Limited to
their spatial uses, these verbs have a reduced element of motion, as in (72)–(74).
(72) En häng-bro som börja-r i botten av bild-en.
DET.INDF rope-bridge COMP.REL begin-PRS in bottom of picture-DET.DEF
‘A rope bridge that begins in the bottom of the picture.’
(73) Une haie qui avance vers la mer.
DET.INDF.Fhedge COMP.REL advance.3SG.PRS toward DET.INDF.Fsea
‘A hedge that advances toward the sea.’
(74) Pen tho yao toò kan oòk ma
be pipe long continue PRON.REFL exit come
chaàk umong.
from cave
‘A long pipe continues out from a cave.’
A third type of NAM-sentence has the opposite character of involving more
information about the type of movement involved. This can be made in different
ways. In this study, we found two such strategies. One is by reference to the
actual motion involved in attaining the particular static situation, as in (75) and
(76), and another way is to use Manner-verbs that convey the velocity associated
with traveling on the Figure, as in (77). This can, of course, be seen as an effect
of the fact that the stimuli only represented linear extensions in space. With a
material also comprised of objects with different shapes and forms, it is possible
The expression of non-actual motion 691
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
that the degree of Manner-information would have been higher (though possibly
still not very high –see Rojo and Valenzuela [2004]).
(75) Cette belle conduite qui semble pénétrer
DET.DEM.Fpretty pipe COMP.REL seem.3SG.PRS penetrate.INF
sous ces roche-s.
under DET.DEM.PL rock-PL
‘This pretty pipe that seems to penetrate under those rocks.’
(76) Pen thang lôt tai umong.
be way penetrate under cave
‘A road penetrates under a mountain.’
(77) Thàňon sen nueng sueng wîng khâw pai
road CLF NUM COMP run enter go
bon phukhǎo.
inside mountain
‘A long road that runs away into a mountain.’
On the basis of (at least) these three different types of NAM-sentences, it is
possible to make further generalizations according to the degree of semantic
information expressing motion. The least specific type includes sentences
expressing Path but not Motion. These are sentences with information only
about the spatial relation construed in somewhat dynamic terms. The clearest
cases were the complex NPs in Swedish with Path adverbs/prepositions. Such
expressions involve specifications of the spatial extension with elements asso-
ciated with motion. I call this Non-actual Path/Direction. NAM-sentences with
generic Motion-verbs or verbs conflating Path and Motion are, in terms of the
degree of motion involved, one step up the hierarchy. In contrast to Non-actual
Path, the spatial relation is not only dynamically conceived but explicitly relies
on motion verbs. A third and final type of NAM-sentences use Motion verbs
expressing more elaborate evocations of motion. This is what may be called Non-
actual Movement. In such sentences, the Figure is described with more elaborate
forms of movements, which can either relate to the shape of trajectory (e.g.,
‘zigzag’,‘snakes’) or the type of movement the entity is associated with, such as
the Thai speakers use of Manner-verbs expressing different velocities dependent
on whether the described object is associated with faster and slower travel. Non-
actual movement would thus not only describe a spatial configuration with
terms of motion, but also include information about state came to be or some-
thing about the motion associated with the extended object in question.
692 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
In sum, the preceding discussion leads us to a three-tiered hierarchy from Non-
actual Path via Non-actual Motion to Non-actual Movement. This corresponds to
an implicational hierarchy, such as those used in typology (Greenberg 1963),
stating that if a language has a particular type of structure then it must also have
all levels to the left of it (see Table 7).
Future research should explore the availability and usage of these three
general types of non-actuality across languages. The present study was not
designed to elicit expressions of Non-actual Movement, and was therefore
unable to provide systematic data in this regard. Due to the scarcity of studies
on NAM-expressions, the research priority was rather to operationalize the
motivations and to implement them in a limited and consistent stimulus mate-
rial designed to elicit NAM-expressions across languages. With the knowledge
that speakers of languages as diverse as Swedish, French and Thai regularly and
spontaneously used NAM-sentences in the present experimental set-up, upcom-
ing studies should preferably include various forms of non-linear configurations
as well as systematically differentiating between entities allowing for fast and
slow travel.
7 Conclusion
In this article, I have explored expressions where motion verbs are used to
describe static configurations in space, what may be called Non-actual motion
sentences. The present study made the distinction between different possible
experiential motivations (based on Blomberg and Zlatev 2014). The three moti-
vations were affordance for motion as directly connected to self-motion, scan-
ning and imagination of motion. These were used to design stimuli for cross-
linguistic investigations. Through an elicitation-based study on Swedish, French
and Thai, it was found that speakers of all three languages use NAM-expressions
prolifically. This supports arguments for the universality and strong cognitive
basis of this linguistic phenomenon.
Table 7: The three types of Non-actual sentences ordered in an implicational hierarchy
Non-actual Path < Non-actual Motion < Non-actual
Movement
Characteristics: No Motion-verb;
Change-of state verbs;
Path-prepositions
Generic Motion-verbs;
Path verbs
Manner-verbs;
Shape-verbs
The expression of non-actual motion 693
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
NAM-sentences were in all three languages most common for stimuli of
affording motion entities represented from a first-person perspective. This result
was interpreted as most strongly in favour of the motivation from self-motion,
but support was also found for the role of imagination and scanning. Since all
motivations were supported to some degree, this speaks for the hybrid nature of
NAM. Future research should add more languages to investigate whether this is
a stable cross-linguistic observation.
A tentative hierarchy ranging from Non-actual Path via Non-actual Motion
to Non-actual Movement was proposed. Where Non-actual Motion expresses
spatial continuity with the help of motion verbs, Non-actual Path signals begin-
ning and endpoints with resources closely connected to motion. Non-actual
Movement adds more motion information pertaining to Manner. That is, if a
language has expressions for the latter, then they also have the other two types
of expressions. With the addition of more languages, the validity of the proposed
hierarchy can be tested.
Even though the three different languages groups behaved similarly, a
semantic analysis of NAM-expressions showed that the resources for expressing
actual motion were retained. The difference between actual and non-actual
motion lies in the fact that the latter strongly relies on semantically bleached
and generic motion verbs. The only exception was the Thai speakers’use of
Manner-verbs to differentiate between the types of velocity associated with
different figure. In sum, NAM-expressions are not only motivated from experi-
ence but are also highly conventionalized with clear language-specific con-
straints on their formation.
Acknowledgements: The research was supported by the Centre for Cognitive
Semiotics, Lund University. I wish to thank Jordan Zlatev for feedback and
comments on various drafts of this paper. Finally, the insightful and construc-
tive comments from three anonymous reviewers greatly improved the paper.
References
Amagawa, Toyoko. 1997. Subjective motion in English and Japanese: A case study of run and
hashiru.Tsukuba English studies 16. 33–50.
Beavers, John, Beth Levin & Shiao Wei Tham. 2010. The typology of motion expressions
revisited. Journal of Linguistics 46. 331–377.
Bergen, Benjamin. 2007. Experimental methods for simulation semantics. In Monika Gonzalez-
Marquez, Irene Mittleberg, Seana Coulson & Michael J. Spivey (eds.), Methods in cognitive
linguistics, 277–301. Amsterdam: John Benjamins.
694 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
Blomberg, Johan. 2014. Motion in language and experience –Actual and non-actual motion in
Swedish,French and Thai. PhD Thesis. Lund University.
Blomberg, Johan & Jordan Zlatev. 2014. Actual and non-actual motion: Why experientialist
semantics needs phenomenology (and vice versa). Phenomenology and the Cognitive
Sciences 13(3). 395–418.
Bohnemeyer, Jürgen, Nicholas J. Enfield, James Essegbey, Iraide Ibarretxe-Antuñano, Sotaru
Kita, Friederike Lüpke & Felix Ameka. 2007. Principles of event segmentation in language:
The case of motion events. Language 83(3). 495–532.
Brandt, Line. 2009. Subjectivity in the act of representing: The case for subjective motion and
change. Phenomenology and the cognitive sciences 8(4). 573–601.
Croft, William, Johanna Barðdal, Willem Hollmann, Violeta Sotirova & Chiaki Taoka. 2010.
Revising Talmy’s typological classification of complex events. In Hans C. Boas (eds.),
Contrastive construction grammar, 201–235. Amsterdam: John Benjamins.
Fillmore, C. J. 1997. Lectures on deixis. Stanford: CLSI Publications.
Gibson, James J. 1977. The theory of affordances. In Robert Shaw & John Bransford (eds.),
Perceiving, acting, and knowing, 127–143. Hoboken: John Wiley & Sons Inc.
Gibson, James J. 1979. The ecological approach to visual perception. Boston: Houghton Mifflin.
Greenberg, Joseph. 1963. Universals of language. Cambridge: Cambridge University Press.
Langacker, Ronald. W. 1987. Foundations of cognitive grammar, Vol. 1:Theoretical
prerequisites. Stanford: Stanford University Press.
Langacker, Ronald. W. 1990. Concept, image, and symbol: The cognitive basis of grammar.
Berlin & New York: Mouton de Gruyter.
Langacker, Ronald. W. 1999. Virtual reality. Studies in the Linguistic Sciences 29(2). 77–103.
Langacker, Ronald. W. 2001. Dynamicity in grammar. Axiomathes 12. 7–33.
Langacker, Ronald. W. 2006. Subjectification, grammaticization, and conceptual archetypes. In
Angeliki Athanasiadou, Costas Canakis & Bert Cornille (eds.), Subjectification: Various
paths to subjectivity,17–41. The Hague: Mouton de Gruyter.
Levelt, Willem J. M. 1996. Perspective taking and ellipsis in spatial description. In Paul Bloom,
Mary A. Peterson, Lynn Nadel & Merrill F. Garrett (eds.), Language and space,77–108.
Cambridge: MIT Press.
Levinson, Steven. C. 2003. Space in language and cognition: Explorations in cognitive diversity,
Cambridge: Cambridge University Press.
Levinson, Steven. C. & David P. Wilkins. 2006. Grammars of space: Explorations in cognitive
diversity. Cambridge: Cambridge University Press.
Matlock, Teenie. 2004a. Fictive motion as cognitive simulation. Memory & Cognition 32(8).
1389–1400.
Matlock, Teenie. 2004b. The conceptual motivation of fictive motion. In Günter Radden &
Klaus-Uwe Panther (eds.), Studies in linguistic motivation, 221–248. Berlin: Mouton de
Gruyter.
Matlock, Teenie. 2010. Abstract motion is no longer abstract. Language & Cognition
2(2). 243–260.
Matsumoto, Yo. 1996. Subjective motion in English and Japanese. Cognitive Linguistics
7(2). 183–226.
Pecher, Diane & Rolf A. Zwaan. 2005. Introduction to Grounding cognition: The role of
perception and action in memory, language, and thinking. In Diane Pecher & Rolf A. Zwaan
(eds.), Grounding cognition: The role of perception and action in memory, language, and
thinking,1–8. Cambridge: Cambridge University Press.
The expression of non-actual motion 695
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
Piaget, Jean & Bärbel Inhelder. 1956. The child’s conception of space. London: Routledge.
Richardson, Daniel & Teenie Matlock. 2007. The integration of figurative language and static
depictions: An eye movement study of fictive motion. Cognition 102(1). 129–138.
Rojo, Ana & Javier Valenzuela. 2004. Fictive motion in English and Spanish. International
Journal of English Studies 3(2). 123–149.
Rojo, Ana & Javier Valenzuela. 2009. Fictive motion in Spanish: Travellable, non-travellable and
path-related manner information. In Javier Valenzuela, Ana Rojo & Cristina Soriano (eds.),
Trends in cognitive linguistics: Theoretical and applied models, 221–239. Frankfurt am
Main: Peter Lang.
Slobin, Dan. I. 2004. The many ways to search for a frog: linguistic typology and the expression
of motion events. In Sven Strömqvist & Ludo Verhoeven (eds.), Relating events in
narrative: Typological and contextual perspectives, 219–257. Mahwah: Lawrence Erlbaum
Associates.
Slobin, Dan I., Iraide Ibarretxe-Antuñano, Anetta Kopecka & Asifa Majid. 2014. Manners of
human gait: A cross-linguistic event-naming study. Cognitive Linguistics 25(4). 701–742.
Stosic, Dejan & Laure Sarda. 2009. The many ways to be located: The expression of fictive
motion in French and Serbian. In Marija Brala Vukanović& Lovorka GruićGrmuŝa (eds.),
Space and time in language and literature,39–60. Newcastle upon Tyne: Cambridge
Scholars Publishing.
Talmy, Leonard. 1983. How language structures space. In Herbert L. Pick . & Linda P. Acredolo
(eds.), Spatial orientation: Theory, research, and application, 225–282. New York: Plenum
Press.
Talmy, Leonard. 2000a. Toward a cognitive semantics, Vol. 1. Cambridge: MIT Press.
Talmy, Leonard. 2000b. Toward a cognitive semantics, Vol. 2. Cambridge: MIT Press.
Zlatev, Jordan. 1997. Situated embodiment: Studies in the emergence of spatial meaning.
Stockholm: Gotab.
Zlatev, Jordan. 2003. Holistic spatial semantics of Thai. In Eugene H. Casad & Gary B. Palmer
(eds.), Cognitive linguistics and Non-Indo-European languages, 305–336. Berlin & New
York: Mouton De Gruyter.
Zlatev, Jordan. 2007. Spatial semantics. In Dirk Geeraerts & Herbert Cuyckens (eds.), The Oxford
handbook of cognitive linguistics, 318–350. Oxford & New York: Oxford University Press.
Zlatev, Jordan & Peerapat Yangklang. 2004. A third way of travel: The place of Thai in motion-
event typology. In Sven Strömqvist & Ludo Verhoeven (eds.), Relating events in narrative:
Typological & contextual perspectives, 159–190. Mahwah: Lawrence Erlbaum Associates.
696 Johan Blomberg
Bereitgestellt von | Technische Universität Berlin
Angemeldet
Heruntergeladen am | 27.10.17 13:54
