-
Epistemic Modality in Greek:
Towards a Constructional and Typological Approach
Accepting language as in some way a result of cognitive
processes, understanding
any language necessarily involves accounting for the ways in
which these processes are
instantiated in that language. The present study represents a
tentative and initiatory
explanation to lay the groundwork for understanding the
conceptualization and construal
of epistemic modality in ancient Greek A constructional approach
can best account for
the primary methods through which epistemic stance is expressed
in Greek texts, and a
typology by which a superior organization and understanding of
primary epistemic
constructions is possible. However, before a typology of
epistemic constructions is
possible, basic components of these constructions, basic
constructions, and some
understanding of their interaction must be analyzed. The purpose
of the present study is
to do just this.
1. Epistemic Modality
It is impossible to proceed with a description of how epistemic
modality is
conceptualized in ancient Greek without first understanding what
epistemic modality is:
Epistemic modality is defined here as (the linguistic expression
of) an
evaluation of the chances that a certain hypothetical state of
affairs under
consideration (or some aspect of it) will occur, is occurring,
or has
occurred in a possible world which serves as the universe of
interpretation
for the evaluation process In other words, epistemic modality
concerns
-
an estimation of the likelihood that (some aspect of) a certain
state of
affairs is/has been/will be true (or false) in the context of
the possible
world under consideration. And this estimation of likelihood is
situated on
a scale (henceforth called the epistemic scale) going from
certainty that
the state of affairs applies, via a neutral or agnostic stance
towards its
occurrence, to certainty that it does not apply, with
intermediary positions
on the positive and the negative sides of the scale.1
Epistemic modality is, then, includes a very large category of
possible statements. There
are any number of methods through which a speaker can construe2
their place on the
epistemic scale. From a pragmatic perspective3, this is not even
limited to word choice.
Intonation, the context of utterance, and other factors make it
possible to turn a simple
indicative statement into a construal of the likelihood of a
situation. For example, the
statement
(1) He was at home.
is, in and of itself, not an expression or construal of
epistemic modality. There is nothing
present in (1) itself to indicate the speakers stance on the
likelihood of the situation
under consideration. It is presented as simple fact. However,
when placed in a specific
context, the same statement can represent the speakers epistemic
stance.
(2) Speaker 1: Theres no possible way he could have been
studying,
because I swear I saw him at the club last night.
Speaker 2: He was at home.
In (2), the statement made in (1) is set in the context of
discourse. A base space4 is
present in which the act of studying by unnamed individual is
under consideration. From
-
this base space a possibility space is constructed by the
Speaker 1 in which the unnamed
individual is in fact not studying, but is at a club. Speaker 2,
in making the same
statement present in (1) is now emphatically denying the state
of affairs5 described in the
possibility space constructed by Speaker 1. In this context, He
was at home is certainly
a construal of epistemic modality because it is now concerned
with the likelihood of a
particular state of affairs. In fact, the very aspects of the
statement in (1) which it
excluded it from the realm of epistemic modality are in (2) the
indication of the speakers
stance on the epistemic scale. In (1), no part of the statement
indicates the possibility that
the state of affairs described did not happen. In (2), this lack
of such indications places
the speaker at the far end (at certainty) on the epistemic
scale, completely rejecting the
possibility of the state of affairs set up by the possibility
space constructed by Speaker 1.
If (1), which contains nothing to indicate the speakers
epistemic stance, can be an
expression of epistemic modality, then the number of ways in
which epistemic modality
can be construed are limitless and form a complex and open
class. Additionally, just as
the lack of any epistemic components can indicate a speakers
epistemic stance, there are
a wide range of epistemic components which may be present in a
statement to locate the
speakers statement somewhere on the epistemic scale: Epistemic
modality can be
expressed by a variety of linguistic forms,such as epistemic
phrases, adverbs, adjectives,
nouns, lexical verbs and participial forms.6 The complexity and
open-ended nature of
the linguistic construal of epistemic modality make it
impossible to capture the infinite
number of ways epistemic statements can be made.
What is possible, however, and what the present study is
concerned with, is the
following:
-
1. What are the typical or proto-typical components involved in
the
general construal of epistemic modality in any given language
(in this
case ancient Greek)?
2. Are these components typically structured in ways which make
possible
meaningful generalizations concerning the nature of
epistemic
constructions7 in that language?
3. If it is possible to identify the basic components and
constructions
generally used, is it possible to construct a typology of the
methods in
which the simpler components and structures are used to indicate
the
speakers epistemic stance?
In other words, simply identifying that any number of words,
types of words, and
combinations of words can express epistemic modality does not
aid in understanding how
speakers in a given language use lexical and syntactical
features in that language in
conceptualizing and construing the likelihood involved in any
given situation. On the
other hand, identifying patterns of lexical and syntactical
constructions, and their
components, used by speakers to express epistemic modality
enables a better
understanding of how speakers use their language to express a
vital aspect of
communication.
In order to lay the groundwork for understanding epistemic
modality in ancient
Greek, several texts of Greek oratory were examined8. From these
texts, a corpus of
epistemic statements were extracted and examined. The typical
components used in these
statements were identified, including lexical items and simple
constructions. From this
examination, the present study constructs a tentative typology
of basic epistemic
-
constructions and identifies not only some important aspects on
the construal of epistemic
stance in complex epistemic constructions, but also notes areas
where further
investigation is necessary.
2. Basic Components
As was already noted, no epistemic markers are necessary in the
construction of
epistemic statements. It is possible for context alone to
transform a statement which lacks
all such markers into an epistemic construction. However, in the
corpus examined, many
common components were used as epistemic markers in epistemic
constructions. Three
of the most important classes of epistemic markers are outlined
below. Excluded are any
components which alone are capable of consideration as epistemic
constructions.
1. Grammatical Mood
Greek grammars9 traditionally recognize a number of moods any
Greek word can take,
including indicative, subjunctive, and optative, which were the
most common in
epistemic constructions. While the indicative mood was the most
common mood used in
epistemic constructions, this likely due to the fact that it was
also the most common mood
used throughout the texts. The optative mood, on the other hand,
while not being enough
by itself to construe epistemic stance, was the only mood
capable of being the primary
component in an epistemic construction in determining epistemic
stance, but even here it
was often combined with an - clause:
(3)
, .
-
For it would be a great stroke of good fortune for the youth, if
one alone
corrupts them, but the others aid. (Plat. Apol. 25b)
(4) ,
.
For if they mean this, I would admit I am an orator, [though]
not in a class
with these. (Plat. Apol. 17b)
Both of the above lines contain a potential optative
construction used as an epistemic
construction. However, the contrast between the two uses shows
the ability of the
potential optative construction to be both the primary component
in an epistemic
construction, as well as a more secondary one. In (3), the
potential optative is the primary
component which is used to construe Socrates epistemic stance.
In (4), the -clause sets
up a possibility space to consider, and the potential optative
construction constructs a
resultative space from the possibility space. The -clause
already indicates that the state
of affairs under consideration is non-actual, and therefore even
before the potential
optative construction Socrates indicates that he doesnt believe
his opponents mean what
he suggests as possible in the possibility space. However, in
(3), the potential optative is
the component of the epistemic construction which initially
construes the non-actuality of
the state of affairs under consideration, and the -clause is
secondary. So while in neither
case, as found in the corpus in general, does the potential
optative construction
encapsulate the epistemic construction, (3) show that it is
quite possible for this
construction to be primary.
2. Tense
-
Various studies have shown the importance of tense in English in
construing the
speakers epistemic stance10. The present study found tense was
also a primary
component in epistemic modality in Greek. Epistemic
constructions in the corpus used all
tenses, and often more than one in the same construction.
However, two general
observations are important. First, future tense indicative
epistemic constructions were less
present. Most epistemic statements used moods as a primary
indicator of the possibility
of future states of affairs. Second, the use of the future tense
indicative tended to place
the speakers epistemic stance towards the certainty end of the
epistemic scale:
(5)
Being cleared of all forms of the charge, neither going into the
holy places
will I profane the sanctity of the gods(Antiph. 2.2.11)
The very fact that the state of affairs under consideration is a
future one makes it
unrealized. However, rather than distancing himself from the
certainty that this
hypothetical state of affairs will be as described (e.g. by
using words such as I think or
by using a different mood/tense of the verb to be) the use of
the future indicative
construes the state of affairs as a certainty..
3. Particles
Greek particles are numerous and are found throughout Greek
literature, used in
all manner of ways. In the corpus examined, the most common use
of particles which
contributed to the construal of epistemic stance were emphatic
(e.g. and )11.
However, other particles were often used as epistemic markers in
interesting ways. For
example, in (4) above, the use of in the -clause without a
corresponding may
-
suggest that some sort of clause to the effect of but they dont
after Socrates states if
they mean this.12 The therefore is an epistemic marker in that
construction,
although this was not a common occurrence in the corpus.
One particle in particular, however, was of paramount importance
in epistemic
constructions. The particle was used in a variety of epistemic
constructions. Just how
this particle should be understood will be discussed in the
section below devoted to .
4. Adjectives/Adverbs
A common element in epistemic constructions in the corpus were
qualifying
words used as secondary components to further determine exactly
where the speakers
epistemic stance was on the epistemic scale. For example, two
common adverbs, and
, were commonly used as components in Simple Epistemic Verbal
Constructions
(SEVCs), but also in Complex Epistemic Constructions (CECs):
(6) ,
If we all were shown to be clever speakers, I know well that
your own
situation would not be in anyway better. (Dem. Ex. 7.2)
(7)
For we know clearly that all of the city is being stained by
him
(Antiph. 2.1.3)
-
In (7), a more complex form of an SEVC (by virtue of other
components being present in
the construction), the adverb simply adds force to the speakers
construal of his
epistemic stance (as does the use of the first person plural,
which serves to pull the jury
into the speakers epistemic construal). In (6), the adverb
serves the exact same
purpose, even though the epistemic construction is far more
complex, using many
elements to construe the speakers epistemic stance.
3. The particle
Any comprehensive grammar of Greek will invariably list a number
of
constructions in which either can or must be used. The variety
of these constructions
make it difficult to define in any meaningful way. Smyth states
explicitly that its
force varies as it modifies the meaning of the moods.13 Goodwin
argues that has, in
general, two separate main uses. In the first of these two it
denotes that the action of
the verb to which it is joined is dependent upon some condition,
expressed or implied.14
In the second is joined with relative and temporal words, and
sometimes final
particles15 rather than verbs. Similarly, Basset divides an into
its appearance in protasis
and apodosis: En grec ancient, la particule avait deux employs
principaux
doublement divergents. Dans un protase, aprs ou , dans une
relative, une
temporelle ou une finale, cest--dire dans des propositions
subordonnes, elle
accompagnait un subjunctive et se liait au mot subordonnantDans
une apodose de
systme hypothtique ou dans une proposition indpendante ou
assimilable une
-
independente, elle accompagnait un optatif ou un temps
secondaire de lindicatif. i16 As
a lexical item, what is it possible to say about the meaning of
, not just in terms of
epistemic modality but in ancient Greek as a whole, given such
diverse usage? First, any
lexical item used with any frequency is almost invariably
polysemous: it has multiple
related meanings that have all been conventionalized to some
degree and therefore there
is no single abstract meaning from which all its uses can be
predicted.17 Rather, an
encyclopedic view18 of the various meanings of will more fully
explain its uses,
and likely enable definitions of the particle that are avoided
in descriptive grammars19.
Second, as noted by Smyth, Goodwin, Schwyzer & Debrunner,
and so forth, the particle
is invariably linked to other words in constructions. As
constructions are not merely
an unstructured list but rather form a structured inventory of a
speakers knowledge of
the conventions of their language.20 These constructions are
represented in a
taxonomic network. In other words, does have meaning, and can be
viewed like any
other lexical item in Greek. It simply is more closely linked
with a limited number of
constructions than other lexical items. To understand the
meaning of , each
construction in which the meaning differs must be understood as
a separate node in the
network of constructions in which is used. Although such an
exercise is beyond the
scope of this study, it would enable an encyclopedic definition
of , and better account
for its use.
i In ancient Greek, the particle has two principle uses for two
different reasons. In a protasis, after or , in a relative, a
temporal or a final, that is to say, in subordinate clauses, it
accompanies a subjunctive and it links itself to a subordinate
word. In an apodosis of [the Greek] hypothetical system or in a
independent clause or something comparable to an independent
clause, it accompanies an optative or a secondary tense of the
indicative.
-
4. The Particle
In addition to being a common component in a variety of
epistemic constructions, the
particle can make a statement an epistemic construction by
itself:
(8)
For I do not suppose those plotting the deaths of their
neighbors devise
and prepare in front of witnesses (Antiph. 1.28)
(8) is an epistemic construction, but is so solely due to the
particle , which alone
acts as an epistemic marker, and which expresses the epistemic
stance of the speaker by
adding at least some doubt to the evaluation of the likelihood
of the state of affairs under
consideration. In this way, functions in a similar fashion to a
SVEC (see below)
with a verb of thinking or supposing.
5. Compositionally Complex Epistemic Words
The fundamental hypothesis of construction grammars is the
continuum between
syntax and lexicon. At the extreme of the lexical side are
atomic morphemes. Although
these can be words, they are often not, and this is particularly
true of Greek, a highly
flectional language. Morphology allows a single word to express
more complexity, and
cannot therefore be considered atomic. Consider the word . From
a constructional
perspective, this word is complex, consisting of an atomic
componential (semantic)
aspect (-) and the morphological aspect which creates
componential complexity by
adding syntactical elements such as number, tense, mood, and
voice. In other words,
can be considered to consist of the atomic component loose whose
morphological
-
properties add the components to form the complex construction I
loose (actively and
indicatively) at this time. Additionally, it is also possible
for a construction composed of
several words to be closer to the atomic side of the
syntax-lexicon continuum by virtue of
be representing a single component. For example, although the
idiom kick the bucket is
composed of several words, it is in fact noncompositional, in
that it the entire expression
is mapped onto die.21
Most words in ancient Greek exist at the level of complex but
bound22 by virtue
being morphologically complex. However, some words exist further
to the syntactic side
of the continuum by way of being schematically conceptualized as
representing,
semantically, more than can be predicted by it morphology.
Simply put, a particular
adverb in Greek, despite its morphological expression and
grammatical class, may be
better interpreted as phrasal: So kann auch ein Adverb statt
eines kausalen oder finalen
oder auch eines anderen Nebensatzes stehenii.23
Among the many possible other Nebenstze and adverb and other
Greek words
are subordinate clauses which construe epistemic stance. These
may be called
Compositionally Complex Epistemic Words (CCEWs). An excellent
example of this
category is the word . Arguably, this word is not typically even
at the level of
complex but bound as it is morphologically invariable (and
therefore, as an
instantiation of a grammatical class it is atomic and
schematic). Pragmatic factors
determine how this word is construed, and often enough it is
used as a simple epistemic
adverb meaning naturally or reasonably. However, at times this
fits very
well into Khners category above, e.g. in Demosthenes speech
Olynthiac (1.10):
ii Thus, an adverb can also stand in place of a causal or final
or also another subordinate clause.
-
(9) ,
Nevertheless, [his] gratitude will be great- as is to be
expected.
A simpler adverbial reading here is inadequate. It fails to
express the phrasal quality of
the adverb.
The question then becomes how these CCEWs are used to convey,
modify, alter,
etc, the speakers epistemic stance. In fact, they are quite
similar to the next type of
epistemic construction considered below in that their primary
function is distancing. In
using a word like , the speaker is offering an epistemic
evaluation, but is doing
so impersonally. In (9), a base space is set up containing the
blessings the gods have
given the Athenians. From this space a possibility space is
built in which any Athenian
can consider these blessings compared to the problems the
Athenians are experiencing. A
resultative space is then constructed in which Demosthenes
determines that any Athenian
will be grateful to the gods despite the current misfortunes.
This resultative space
represents a positive epistemic stance formed from a Complex
Epistemic Construction,
and will therefore be considered in full below. Here, what is
important is how the CCEW
functions in the resultative space. This entire space is built
from a possibility
space. In it, is used to construe Demosthenes epistemic stance
on the
probability that any given Athenian, having weighed the
blessings of the gods against the
current problems, will be grateful to the gods. However, does
not directly
connect this stance to Demosthenes. Even though it, in fact,
merely expresses
Demosthenes evaluation of a likelihood, it is designed to appear
that this is not an
opinion of Demosthenes at all. Demosthenes does not say I expect
X, in which it
-
would be clear that Demosthenes alone considers X as likely.
Instead, he uses
as a way of construing the hypothetical situation as likely from
any perspective.
The function of CCEWs is to do just that: construe an epistemic
stance, a personal
evaluation of likelihood, as a general truth.
6. Simple Epistemic Vebal Constructions (SEVCs)
The simplest of the simple epistemic constructions in Gree, are
phrases or clauses which
use a first person verb (or participle) of thinking, perceiving,
believing, etc. Examples
found in the corpus include , , , , , and .
In a Simple Epistemic Verbal Construction (SEVC), the place of
the speakers statement
on the epistemic scale, in the indicative mood at least, is
determined solely by the
semantics of the verb selected. This is hardly surprising, as it
is fairly obvious that an
epistemic stance taken with the statement I believe does not
have the force of one taken
with I know. Consider, for example, the difference between the
following:
(9) ,
.
And I know indeed from earlier reports already it has happened
that sometimes
those murdered, and other times those having murdered, have not
been found out.
(Antiph. 5.67)
(10)
.
For you will not, I suppose, deny what you did in front of all
the Athenians.
(Lys. 13.32)
-
It is true that the particle attached to the first person
pronoun, as well as the pronoun
itself, in (9) add emphasis apart from the verb choice itself24.
However, even without
these the use of in (9) versus the use of in (10) are enough to
place the
two statements at different points on the epistemic scale. The
former makes a much
stronger epistemic stance, and the statement is made as if there
is no doubt at all, and can
be no doubt, that the past events in question are as the speaker
asserts. Examining the
second statement from a pragmatic point of view, it is easy to
see why the speaker would
not wish to commit himself as in (9). On the one hand, he is
asserting his belief that his
opponent will do X. However, as doing X would actually be an
honest action on the part
of his opponent, a stronger epistemic stance is undesirable, as
it would be the equivalent
of a commitment to the honest nature of his opponent.
Pragmatics is a central factor which often dictates the choice
of verb in SEVCs.
What is most important here is that in these simple
constructions it is the verb (or verbal
variant such as a participle) which makes the construction an
epistemic one. As such,
these constructions are actually somewhat rare. Although
included in this section for
contrastive purposes, the statement (9) above is actually a
Complex Epistemic Statement
due to the use of an emphatic pronoun and particle. It uses
other components than the
verb alone to make the statement an epistemic construction. In
the corpus examined, this
was far more the norm than the exception.
7. Impersonal Epistemic Constructions
Impersonal Epistemic Constructions (IECs) function similarly to
CCEWs
described above. These constructions can be subdivided into
three different types.
Type 1: Qualifier (Q) + 3rd Person Copula
-
As in English and many other languages, (e.g. French and German)
Greek allows
impersonal constructions by combining a qualifier (adjective,
participle, etc) with a form
of the third person copula. A few such qualifiers are commonly
used in IECs, such as
, , , etc. For example:
(11)
It is not [at all] unclear that [the defence] themselves were
avoiding clear
knowledge of the facts (Antiph. 1.13)
In (11), as is common in such constructions25, the copula has
been omitted and must be
understood. The epistemic component of the above construction is
the adjective (here
used adverbially) . The pragmatic function of such a
construction is again to
make the epistemic stance of the speaker appear to be fact
rather than opinion.
Type 2: Distanced Epistemic Impersonal Verbs
Greek has a number of verbs which are commonly used. In the
examined corpus,
one of the most common was deontic modality (e.g. using ).
However, possibility
that a particular action, situation, etc. could be done/would
happen/etc. was also common
(generally using the impersonal verb ). However, this
possibility should not be
confused with epistemic modality, as it concerns the ability of
a particular state of affairs
to occur, rather than likelihood. In fact, there was not a
single instance in the entire
corpus of this type of construction. This was a curious finding,
given the prevalence of
such constructions in other languages (e.g. It is likely that X
is Y).
Type 3. Personalized Impersonal Epistemic Constructions
-
Despite the above, there were numerous examples in the corpus
examined of
impersonal verbs used in epistemic constructions but that
brought the speaker himself
into the evaluation using the first person pronoun in the dative
case (). The most
common verb used in this construction was :
(12)
That I speak the truth has been attested, and it appears quite
evident to me [that]
this is the opposite of what the prosecution alleged.
Although this epistemic construction is formed impersonally, the
use of the first person
pronoun connects the speaker with the evaluation. The purpose of
this construction is not
to distance the speaker from the evaluation, but is rather a
matter of construal and
perspectivization. By using this epistemic construction, rather
than stating something to
the effect of I believe it is quite evident that. a mental space
is created in which the
focal point of attention is on the situation (in which the
prosecution is shown to be
wrong) rather than the epistemic stance of the speaker. In other
words, this type of IEC
allows the speaker to make an epistemic stance, connect
themselves with that stance, and
yet ensure that another facet of the discourse is the focus of
attention.
8. Complex Epistemic Constructions
1. Conditions
Unlike epistemic modality, the topic of conditions in ancient
Greek has received a
great deal of attention26. Additionally, the study of the nature
of conditionals themselves
has been a focus of intense study. Even what constitutes a
condition is debated27. Given
all of the above, attempting to go into Greek conditionals in
any detail would either be a
-
waste of time, if nothing new was put forward, or would require
a study several times the
size and complexity of the present one, in order to
appropriately address the mass of
literature already published on conditionals in general and
Greek conditionals
specifically.
What is important to note, however, is that all conditionals in
Greek which may
be considered epistemic constructions are complex, and use a
variety of basic epistemic
components (e.g. verbal mood, the particle , verbal tense, etc).
Additionally, most
grammars of ancient Greek deal with conditionals as relatively
fixed combinations of
these elements. This is largely appropriate, as there are set
constructions which for
specific conditionals (e.g. a so-called irrealis conditional in
the present is constructed
using + the imperfect indicative in the protasis, and aorist or
imperfect + in the
apodosis). What has received no significant attention is the
other components usually
involved in conditions. In the corpus examined, rarely were
epistemic conditional
constructions formed using only the components described in
typical grammars of Greek.
As far as conditionals are concerned, what will be examined
below are how components
not required by particular conditional constructions are used in
certain CECs, as well
as how simpler epistemic constructions are used in conditionals.
Two conditionals, one
using a typical construction described in any Greek grammar, and
the other an atypical
construction, are examined below with these purposes in
mind.
(13) ,
If we all were shown to be clever speakers, I know well that
your own
situation would not be in anyway better. (Dem. Ex. 7.2)
-
This type of conditional is often referred to as future less
vivid or something similar.
The central function of this type of condition is to set up an
unlikely hypothetical state of
affairs, and infer from that possibility space28. What is
important to understand is that the
less vivid aspect is all contained within the protasis. If that
hypothetical state of affairs
is realized, the use of this construction does not make it less
likely, in the speakers mind,
that the apodosis will be realized. Hence, the apodosis of this
conditional construction not
only combines a SVEC into the entire construction, and adverbial
epistemic component
() is also used to make it clear that, in the unlikely event the
protasis is realized, the
apodosis will surely be as well.
(14) ,
.
If indeed I have entered court [at some point](followed by a
number of
other hypothetical situations) Let the one wishing rise up now
and
refute me! (Andoc. 1.23)
In (14) Andocides begins by setting up a possibility space in
which several states of
affairs (all having to do with things which might implicate him)
are considered. The first
-clause sets up this possibility space, but there are no
epistemic markers in any of the
various connected clauses which are used in the construal of
Andocides epistemic
stance. In fact, there are no direct epistemic markers (e.g. a
word or particle with a
particular semantic value used to express his epistemic stance,
or a verbal mood/tense
combination which may do the same) anywhere in the entire
condition. Andocides
epistemic stance is made indirectly by a third person imperative
(). By issuing
a command to his audience to refute him, the implication is that
there is no possible
-
refutation, because none of the various states of affairs ever
happened. Once again, taking
pragmatics into consideration is necessary to understand the
construal of epistemic
modality.
These are only two example of ways in which various standard
components of
Greek epistemic constructions (i.e. SEVCs and verbal moods) can
interact with Greek
conditionals in the construal of epistemic modality. Far more
methods were found in the
corpus. In fact, every single component discusses above was
found in conditionals in the
corpus. Although it is beyond the scope of the present study, a
more thorough
investigation of epistemic markers in the various Greek
conditionals is necessary.
2. Other CECs
Before going on to make some generalizations about CECs in
Greek, and some
points for further investigation, it is useful to look at
examples outside of conditionals.
After all, epistemic modality in Greek is most frequently
expressed in CECs, and this is
hardly limited to conditionals.
(15) , ,
,
.
It would be fitting, gentlemen, whenever someone willingly
brings
themselves into the dangers [of court], for you to have they
exact same
opinion they of them as they themselves have. (Andoc. 1.3)
In (3), the hypothetical state of affairs (i.e. the unrealized
aspect of the situation under
consideration), is constructed by the use of the optative + in a
potential optative
construction. What is interesting about the above construction
is that, despite the phrasal
-
use of to note the suitability of hypothetical state of affairs
constructed, as well
as the use of the particle to add force to the statement, from
an epistemic standpoint
there is little in the construction in which the speaker
expresses his epistemic stance. In
other words, although this is a Complex Epistemic Construction,
the epistemic modality
aspect is secondary. Pragmatic factors explain this secondary
aspect. The potential
optative is used impersonally (like ) to distance the speaker
from the state of
affairs, not only from an epistemic perspective but also from
the truth of the statement
altogether. The likelihood that the state of affairs will
actually occur is only realized as
possible by the potential optative construction, and nothing
else in (15) construes the
speakers view on how possible or likely that state of affairs
is. The purpose for the
distancing in epistemic stance is secondary. The force of the
entire statement is not on the
likelihood that the state of affairs will be realized, but
rather to convince the gentlemen
of the jury that this state of affairs should be conceptualized
by all in a particular way (i.e.
as fitting or perfectly reasonable). By distancing himself from
this conceptualization,
the speaker makes his opinion appear to be fact. If the speaker
had made his epistemic
more personalized, it would have taken away from the illusion
that the statement is not
opinion at all.
(16)
All of those present would have spoken better than but a few of
them
concerning [the poetry] they themselves had made. (Plat. Apol.
22b)
-
Socrates in discussing the failure of the poets to understand
their own work, constructs a
possibility space in which all of those present were also
present with him at the time
when he was addressing the poets. The imperfect tense, combined
with the particle , is
used to construct the speakers epistemic stance: this
hypothetical state of affairs will not
be realized. The potentiality is constructed from the particle ,
while the speakers
epistemic stance is construed both by and by the verbal tense
both of
(imperfect) and (pluperfect). In other words, the use of past
tense verbs
places the time of the hypothetical state of affairs in the
past. As it is in the past, by
necessity it either happened or didnt. Because makes the
statement one of potential,
clearly the hypothetical state of affairs did not in fact occur.
The past tense makes the
mental space constructed either hypothetical or not, and if it
is hypothetical then it is
unrealized, and the epistemic stance of the speaker is clear (on
the certainty end of the
epistemic scale that this hypothetical state of affairs will
never happen).
9. General considerations
From the examination of epistemic constructions in the corpus, a
number of
important general observations were made, as well as several
notes of interest where
further investigation is needed.
1. As noted above, hypothetical state of affairs constructed
from past tense verbs
almost invariably construed epistemic modality on the certainty
end of the
epistemic scale. Although it is quite possible to speak of past
events as possibly
happening when the speaker is unsure, this rarely happened in
the corpus
examined.
-
2. Almost all epistemic constructions were complex. That is,
quite rarely were
simple epistemic constructions used to construe epistemic
stance. Most epistemic
constructions were either conditionals or simple epistemic
constructions which
used particles, adverbs, adjectives, and other components to add
to the construal
of epistemic stance.
3. Particles of emphasis were used everywhere as components in
the construal of
epistemic modality. Additionally, the particles and were of
great
importance in many epistemic constructions. However, particles
were present in
almost all epistemic constructions in the corpus, and of
singular difficulty was the
determination of when these particles were used as epistemic
markers, and when
they were used for other purposes. A focus on particles from a
standpoint of
epistemic modality alone would therefore be uniquely helpful in
understanding
epistemic modality in Greek.
4. Epistemic constructions in Greek may be broadly divided into
impersonal and
personal categories. Further, the more personalized the
epistemic stance was,
the more it was possible for the speaker to place himself
anywhere on the
epistemic scale. Impersonal epistemic constructions were often
used as distancing
mechanisms.
5. In general, no singular component or simple epistemic
construction stood out as
the primary method of construal of epistemic modality. However,
the importance
of various components or constructions did stand out when the
place on the
speaker on the epistemic scale is considered. For example, the
further from
-
certainty the speaker was on the epistemic scale, the more
verbal mood was
important, and other components played a secondary role and were
less prevalent.
6. Given its prevalence in epistemic constructions, a typology
of -constructions
from an epistemic standpoint would be useful not only in gaining
a better
understanding of the encyclopedic meaning of the particle, but
in constructing a
typology of epistemic constructions in general.
10. Conclusions
There is far more to the construal of epistemic modality than is
covered above. The
topic itself is complex, and also forms an open-ended class. The
present study was
concerned only with laying the groundwork towards a typology of
epistemic
constructions. This was accomplished by identifying the
components and simple
epistemic constructions used in the construal of epistemic
modality. Additionally, several
epistemic constructions were analyzed so that points of interest
could be identified for
further study. Furthermore, several Complex Epistemic
Constructions were examined,
and again analyzed to identify points of interest. Finally, some
general observations were
made. From all of the above, it is clear that although a great
deal of investigation is
necessary towards the goal of understanding epistemic modality
and creating a typology
of epistemic constructions in Greek, various regularities in
expression appear to make
this goal possible.
1 Nyuts (2001), pp. 21-22.
-
2 The terms construe and construal are used here as they are in
research in cognitive
linguistics: The cover term that has come to be used for
different ways of viewing a
particular situation is construal. (Verhagen, 2007, p. 38). In
other words, the construal
of epistemic modality refers to the ways in which the speaker,
expresses, represents,
captures, etc, her or his particular viewpoint of this
modality.
3 One outcome of research in functionalist linguistics is the
importance of the context in
which speech or language occurs. Pragmatic as used here relates
to the observation that,
in the context of the human behavioral repertoire, language has
a specific role to play,
viz. (primarily) to allow communication with other members of
the speciesHence,
investigating language also unavoidably means accounting for how
this system fulfills
this communicative function (Nyuts, 2001, p. 22)
4 The term space here, e.g base space, hypothetical space, etc,
is borrowed from
Fauconnier (1985; 2007), who describes mental spaces as a basic
and common aspect
of human language.. The concept is used in a number of different
approaches to language
in cognitive linguistics. It refers to very partial assemblies
constructed as we think and
talk for purposes of local understanding and action (Fauconnier,
2007, 351). In other
words, during the course of discourse, speakers construct local
spaces where important
elements of the conversation are located and may be expanded
upon. It would be helpful
to condiser and example: Suppose that we are engaged in a
conversation about Romeo
and Juliet and the following statement is made:
(1) Maybe Romeo is in love with Juliet.
The English sentence brings in a frame from our prestructured
background cultural
knowledge, X IN LOVE WITH Y, with two roles highlighted (the
lover x and the loved
-
one y)The word maybe is a space builder; it sets up a
Possibility space relative to the
discourse Base space at that point. The Base space contains
elements a and b associated
with the names Romeo and Juliet (Fauconnier, 2007, 355). In the
construal of
epistemic modality in Greek, a possibility space is invariably
constructed out of the base
space, and the epistemic stance of the speaker concerns the
likelihood of the realization
of the state of affairs in that possibility space. The term
state of affairs is borrowed from
Functional Grammar and is an all-encompassing term covering the
entity to which the
whole of the predicate with its arguments refers, irrespective
of whether this entity is an
event, an action, a situation, etc (Wakker, 1994, p. 7n14).
5 See note 4 above.
6 Krkkinen (2003) p. 32.
7 The term constructions here refers to constructions as they
are conceived of in various
construction grammars, which posit that there is a uniform
representation of all
grammatical knowledge in the speakers mind in the form of
generalized construction
everything from words to the most general syntactic and semantic
rules can be
represented as constructions (Croft, 2007, p. 471). The most
important aspect of
construction grammars is the idea that syntax and lexicon form a
continuum, rather than
being two separate components (Goldberg, 1995, p. 7, Langacker,
2008, p. 161).
Additionally, constructions are organized in a taxonomic network
(Croft, p. 476-477),
and vary from very fixed to highly variable and schematic (ibid,
pp. 470-471). All of
speech is formed from combining various constructions
(Langacker, 2008, chaps. 1 and
6).
-
8 All of the texts were taken from the Perseus Project of Tufts
University
(http://www.perseus.tufts.edu/hopper/collection?collection=Perseus:collection:Greco-
Roman). These included the speeches of Antiphon, Andocides,
Demosthenes, Lysias, and
the Apology of Plato.
9 The Greek grammars consulted in the present study include
Smyth (1956), Humbert
(1960), Schwyzer & Debrunner (1966), Goodwin (1889), and
Khner & Gerth (1904).
10 See, e.g. Dancygier & Sweester (2005, sect. 2.4) and
Dancygier (2004, chap. 2).
11 On the usage of particles in general, as well as the emphatic
usage of these particles,
see Denniston (1950).
12 See Smyth, sect. 2896.
13 Sect 1762.
14 p. 64.
15 Ibid.
16 p. 27.
17 Langacker, pp. 37-38.
18 As opposed to a dictionary view (see Langacker, pp. 38-39).
An encyclopedic view of
lexical semantics, lexical meaning resides in a particular way
of accessing an open
ended body of knowledge pertaining to a cerain type of entity
(Langacker, p. 39).
19 In other words, rather than simply describe a force has, it
is possible to describe
its meanings as if it were any other lexical item. This is true
of all particles, and would be
a superior method of describing them in Greek grammars.
20 Croft, pp. 476-77.
21 Croft, pp. 467; Croft and Cruse (2004) p. 253.
-
22 The levels of the syntax-lexicon continuum may be found in
Croft and Cruse, p. 255:
23 Khner & Gerth, p. 115.
24 Goodwin pp. 115-116.
25 Schwyzer & Debrunner p. 623, Smyth sect. 944.
26 Apart from Greek grammars, see, e.g., Wakker, Seiler (1997),
and Greenberg (1986).
27 Wierzbicka (1997) pp. 15-18.
28 Goodwin, pp. 168-69.