See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226722839

From the Grammaticalization of Viewpoint Aspect to the Grammaticalization

of Situation Aspect

Chapter · November 2007

DOI: 10.1007/978-1-4020-4755-8_7

CITATIONS READS

18 116

1 author:

Katalin É. Kiss
Hungarian Academy of Sciences
53 PUBLICATIONS 2,125 CITATIONS

SEE PROFILE

All content following this page was uploaded by Katalin É. Kiss on 14 July 2016.

The user has requested enhancement of the downloaded file.

Chapter 7
From the grammaticalization of viewpoint aspect to the gramaticalization of situation
aspect
Katalin É. Kiss

I. Introduction
As illustrated in chapter (1), in present-day Hungarian basically all accomplishment and
achievement predicates denoting a delimited change of state or change of location have a
verbal particle. In Standard Hungarian, there is still a handful of verbs left which are
ambiguous with respect to situation aspect; they can be interpreted both as atelic process
predicates and as telic accomplishment or achievement predicates. In spoken Hungarian,
however, these verbs have also developed a clearly telic variant associated with a verbal
particle. For example:

(1)a. Standard Hungarian:

A kuratórium értékeli a pályázatokat.
the council evaluates the applications

b. Spoken Hungarian:
A kuratórium ki -értékeli a pályázatokat.
the Council out evaluates the applications

(2)a. Standard Hungrian:

Az orvos fertőtlenítette a sebet.
the doctor sterilized the wound

b. Spoken Hungarian:
Az orvos le -fertőtlenítette a sebet.
the doctor down sterilized the wound

(3)a. Standard Hungarian:

Ellenőrzöm a megadott telefonszámot.
check-I the given phone-number
‘I check the given phone number.’
 b. Spoken Hungarian:
Le -ellenőrzöm a megadott telefonszámot.
down check-I the given phone-number

In Spoken Hungarian, Latin prefixed verbs are also supplied with a Hungarian particle, for
example: le-degradál ’down-degrade’, át-transzformál ’through-transform’, el-deformál ’off-
deform’, ki-disszidál ’out-defect’. Compare:

(5)a. Standard Hungarian:

Az autó ajtaja deformálódott az ütközéskor.
the car’s door deformed the crash-at
‘The door of the car deformed at the crash.’

b. Spoken Hungarian:
Az autó ajtaja el -deformálódott az ütközéskor.
the car’s door off deformed the crash-at

This kind of non-standard usage of the verbal particle is also creeping into literary Hungarian.
Observe, for example, the following citation from a 2004 poem by Á. Nádasdy:

(6) ”Minek a szeretet,

what-for the love
’What is love for

ha nem hunyok szemet, ha gonoszul

if not shut-I eye-ACC if wickedly
if I don’t shut my eyes, if wickedly

ki -ábrázolom, kinek mije van?”

out represent-I who-to what is
I represent (out) who has what?’
Ádám Nádasdy: „Az ábrázolás” ‘The Representation’ (Élet és Irodalom July 23, 2004)
In Standard Hungarian, codified e.g. in Hungarian dictionaries, ábrázol ‘represent’ has no
verbal particle associated with it, nevertheless, the intended telicity of the predicate forces the
poet to supplement it with a particle.
Standard Hungarian, which is based primarily on the classical literature of the 19th
century and the first half of the 20th century, is more conservative than Spoken Hungarian.
The slight discrepancy between Standard Hungarian and Spoken Hungarian with respect to
telicity marking suggests that the grammaticalization of telicity by means of a resultative or
terminative particle has been a historical process which has basically been completed in the
spoken language, and is nearing to its completion in the more conservative standard language.
The gradual grammaticalization of situation aspect marking, i.e., the extension of telicity
marking to all sentences describing a delimited change of state or a delimited change of
location, has spanned about 800 years in the history of Hungarian. The first surviving
coherent Hungarian text, Halotti beszéd és könyörgés ‘Funeral speech and prayer’, written
between 1192-95, contains only a single instance of the verbal particle. In the Old Hungarian
langauge of those times, aspect was marked by a rich system of verbal inflection, including
both a viewpoint aspect morpheme, and a tense morpheme. The emergence of telicity marking
verbal particles, i.e., the grammaticalization of situation aspect, was soon followed by the
simplification of verbal inflection, i.e., by the disappearance of viewpoint aspect marking.
This chapter will raise the question if the near simultaneity of these two processes was a mere
coincidence, or perhaps the emergence of situation aspect marking made the morphological
marking of viewpoint aspect redundant. It will argue that viewpoint aspect can in the great
majority of cases be inferred from situation aspect – thereby weakening the claim put forth
e.g. by Smith (1991) that situation aspect and viewpoint aspect represent independent
subsystems of aspect.
The chapter is organized as follows. Section 2 describes the aspectual system of Old
Hungarian, analyzing its complex inflection as a combination of viewpoint aspect marking
and tense marking. Section 3 provides a morphosyntactic representation of the facts described
in Section 2, employing the framework of Demirdache and Uribe-Etxebarria (2004). Section 4
briefly surveys the stages of the emergence of telicity marking, and of the gradual
disappearance of viewpoint aspect marking. Section 5 argues that the overt morphological
marking of both types of aspect is redundant because viewpoint aspect can be inferred from
situation aspect.

2. The encoding of tense and viewpoint aspect in Old Hungarian

Whereas present-day Hungarian only has two tenses, an unmarked present tense, and a past
tense marked by -t/tt, Old Hungarian had a variety of verb forms denoting past events. A verb
stem like mond ‘say’ occurred in the following temporal–aspectual versions: mond
‘say.PRES.3SG’, mond-a, mondo-tt, mond vala, mondott vala. The analysis of the contexts in
which these verb forms occur will lead us to the conclusion that they had the following
temporal–aspectual values:

(6) mond – simple present

mondott – present perfect
monda – simple past
mond vala – past imperfective
mondott vala – past perfect

The simple present was used to describe states and processes going on at the time of the
utterance. Observe the following examples, quoted in present-day spelling:

(7) Látjátuk feleim szümtükhel, mik vogymuk?

see-you fellows-my eyes-your-with what are-we
‘Do you see, my fellows, with your eyes what we are?’
(Halotti beszéd és könyörgés ‘Funeral Speech and Prayer’, 1192-95, in Forgács et
al. (1996b:7))

The -t/tt suffix, which is identical with the past tense suffix (as well as the past participle
suffix) of present-day Hungarian, has been claimed in traditional literature to represent a past
tense suffix used in personal communication. Whereas in medieval Hungarian literature, the
general tense of story telling is the simple past marked by -a/e, there is often a switch to the -
t-marked tense when the story contains a direct quotation. Observe, for example, the
following section of the Vienna Codex (1450):

(8) Orpha megapolá ő napát s meg -fordola.

Orpha kissed her mother-in-law and around-turn-PAST-3SG.
‘Orpha kissed her mother-in-law, and turned around.

Rut eggyesőle ő napával. Kinek mondá Noémi:

 Ruth joined her mother-in-law whom said Naomi
Ruth joined her mother-in-law. Naomi told her:

Ime te rokonod meg -fordolt.

look your relative around-turn-PERF-3SG
Look, your relative has turned around.’
(Bécsi-kódex (1450:2), quoted by E. Abaffy (1992:154))

In the following quotation from a repentant psalm, the speaker, talking to the Lord in first
person singular, uses only -t/tt-marked verbs:

(9) Jaj nekem, mert haragossá én ellenem tött-em én megváltómat...

woe to.me because angry me against make-PERF-1SG my savior
‘Woe be to me because I have made my savior angry against me!’

Igaz utat elhagytam, és az járatlan utakon messze...forgott-am.

true way-ACC leave-PERF-1SG and the untrodden ways-on far move-PERF-1SG
‘I have left the true way, and I have moved far on untrodden ways.’
(Festetich-kódex (1493:396), cited by E. Abaffy (1992:155))

As illustrated by these examples, the sentences containing a verb supplied with a -t/tt suffix
”present a state of affairs with characteristics due to the prior situation” – which is the
description of present perfect tense by Smith (1991:147). In such sentences, the speaker
focusses on the resultant state of a past event, which is still in effect at the time of the
utterance. In other words, a -t/tt marked verb form describes a past event seen from a present
viewpoint. The aspect of these verbs, establishing a relation between a past event and a
present reference time, is the perfective, and the tense, establishing a relation between a
present reference time and a present utterance time, is the present.
The illusion that the -t/tt-marked tense is a past tense used in personal communication
arises because present perfect involves a reference time including the utterance time, which is
typical of personal communication, e.g. of story-telling by an affected party. Biblical or
historical story-telling, i.e., story-telling by a non-participating party, rarely uses a reference
time including the utterance time; such stories rarely have a result state extending to the time
of the story-telling.
 The -a/e-marked past tense appearing in some of the sentences quoted above, called
’historical past’, represents the Old Hungarian simple past. It was used to describe past
events in the case of which the event time and the reference time are identical; in other words,
the past is not seen as related to the present. This is the tense of story-telling by a non-
participant, e.g.:

(10) Száll-a alá poklokra.

descend-PAST.3SG down hells-into
‘He descended into hell.

Harmadnapon halottaiból feltámad-a.

third-day-on dead-his-from resurrect-PAST.3SG
The third day He arose again from the dead.

Felmén-e a mennyekbe…
ascend-PAST.3SG the heavens-into
He ascended into heaven’
(Apostle’s Creed)

Crucially, this is the tense used in personal communication, as well, if the sentence contains a
time adverbial referring to a past point of time or a past period. In such sentences, the time
adverbial specifies the reference time:

(11) Enyingen ír -á -m ez levelet pénteken

Enying-at write-PAST-1SG this letter Friday-on
‘I wrote this letter at Enying on Friday’
(a letter from 1529, quoted from Magyar Nyelv (37:277) by E. Abaffy (1992:153))

The verb form mond vala represents the past imperfective. It was used, on the one hand, in
cases when the reference time, i.e., the viewpoint, is internal to the time span of a past event.
This is the context triggering the use of past imperfective e.g. in the following quotation from
a fable of Száz fabula ‘Hundred fables’ by Gáspár Heltai (1566):

(12) Melléje gyűlvén az mezői egerek, ott játszadoznak vala környüle,

 near-him gathering the field mice, there play-IMPERF.3PL be-PAST around.him
‘Having gathered around him, the field mice were playing around him,

és a kergetésbe az egyik reája szökellék az oroszlánra.

and the chase-in the one on.him hopped the lion-on
and during the chase one of them hopped on the lion.’
(Gáspár Heltai: 16. fabula: Az oroszlánról és az egérről ‘On the lion and the
mouse’, quoted by Forgách et al. (1996b:10))

The reference time is specified by the event time of the predicate reája szökellék ‘hopped on
him’. The playing event denoted by játszadoznak vala ‘were playing’, taking place in the past
(prior to the utterance time), is going on before, during, and after the reference time.
Another function of the past imperfective was to mark past habitual events. Observe, for
example, the following quotation from the legend of Saint Margit. The sentences which
describe Saint Margit’s usual activities all contain a verb in the past imperfective. On the
other hand, when the author, Lea Ráskai, tells a unique event in the life of Saint Margit, she
switches to the simple (historical) past:

(13) Ez nemességes szent szűz magyari királnak nemes leánya hetet

this noble saint maid Hungarian king’s noble daughter week-ACC

tart val-a, az konyhán főz val-a az sororok-nak,

keep-IMPERF.3SG be-PAST, the kitchen-in cook-IMPERF.3SG be-PAST the sisters-DAT

fazekat mos val-a, tálakat mos val-a,

pot-ACC wash-IMPERF.3SG be-PAST, bowls-ACC wash-IMPERF.3SG be-PAST,

az halakat megfaragja val-a … Azért mikoron szent Margit asszony

the fishes-ACC carve-IMPERF.3SG be-PAST therefore when saint Margit lady

egy napon szokása szerént az mosdóvíznek moslékát akarná kivinni

one day her.habit according the washing.water’s leavings-ACC want-COND to.take.out

az refektóriumból, de nem vi-het-é el az víznek sokaságáért,

 the refectory-from but not take-POT-PAST.3SG off the water’s mass-for,

hív-a hozzá egy sorort…

call-PAST.3SG to.it a sister

‘This noble saint maid, the Hungarian king’s noble daughter, would have turns on duty,
she would cook in the kitchen for the sisters, she would wash pots, she would wash bowls,
she would carve the fish… When one day lady Saint Margit wanted to take out the
dirty leavings of the water for washing, but could not take it because of the mass of the
water, she called a sister to it…’
(Margit legend 1510, in Forgács et al. (1996b:8))

The past imperfective could also be combined with an achievement verb marked by a verbal
particle, in which case it expressed that the event did not culminate; it stopped after a
preparatory phase. E.g.

(14) Es oz gyimilcsnek úl keseröü vola vize, hugy turkokat

and that fruit-DAT so bitter was water-its that throat-their-ACC
migé szokosztja vol-a.
PRT burst-IMPERF.3SG be-PAST
‘And that fruit had such a bitter juice that it was bursting their throats.’
(Halotti beszéd és könyörgés ‘Funeral Speech and Prayer’ 1192-95, in Forgács et
al. (1996b:7))

This sentence relates the story of Adam and Eve in Paradise, therefore we know that the apple
which „was bursting their throats” (migé szokosztja vola) did not actually burst their throats,
i.e., it did not kill them, merely it was about to kill them.
Let us assume that habitual activities, e.g. those illustrated in (16), form an extended
event. Then it is true for all three types of contexts in which the imperfective past occurs that
the reference time (i.e., the viewpoint) is internal to the event time, and the utterance time
follows the reference time.
The verb form mondott vala represents the past perfect tense. It was used to describe past
situations from a reference point which followed the event time but preceded the utterence
time. For example:
(15)a. És megemlékez-é-k Péter az igéről, kit mondo-tt val-a nekik
and commemorate-PAST-3SG Peter the word-about that say-PERF.3SG be-PAST them
‘And Peter commemorated the word that he had told them.’
(Müncheni-kódex (1466:103), quoted by Bánhidi 1941:24)

b. Meglel-t-em én juhomat, ki elvesze-tt val-a

find-PERF-1SG my sheep that astray.go-PERF.3SG be -PAST
‘I have found my sheep that had gone astray.’
(Müncheni-kódex (1466:146), quoted by Bánhidi (1941:24))

In both sentences, the time of the event described in the first clause represents the reference
time for the verb of the second clause. The verb of the second clause denotes an event that
took place prior to the reference time, hence its aspect is perfective. The utterance time
follows the reference time, therefore the tense of this verb is past.

3. The syntactic representation of Old Hungarian tense and viewpoint aspect

The current standard representations of complex tense–aspect systems (e.g. Hornstein 1993)
go back to the theory of Reichenbach (1947), which derives the different English tenses
(incorporating also aspect) from the varying relations of speech time, reference time, and
event time. From among the neo-Reichenbachian theories, I have chosen the approach of
Demirdache and Uribe-Etxebarria (2004) as the framework of representing Old Hungarian
tenses because of the clear and plausible syntactic claims that it makes.
In the framework elaborated by Demirdache and Uribe-Etxebarria (2004), Tense and
Aspect are spatiotemporal predicates projecting a temporal argument structure in syntax.
AspP immediately dominates VP, and TP immediately dominates AspP. Event Time is a
temporal argument adjoined to VP; Reference Time (i.e., the viewpoint in the terminology of
Smith (1991)) appears in the specifier of AspP; and Utterance Time is in the specifier of TP.
Aspect establishes a ‘within’ or ‘after’ relation between Reference Time and Event Time; and
Tense establishes a ‘within’, ‘after’, or ‘before’ relation between Utterance Time and
Reference Time. In the case of simple tenses, Event Time and Reference Time are identical.
In this framework, simple present can be represented as follows:

(16) Simple Present

 TP
/ \
UT T’
/ \
T AspP
within / \
RTi Asp’
/ \
Asp VP
within / \
ETi VP……

Utterance Time is within the coindexed Reference Time/Event Time; i.e., the Event Time
(and the Reference Time which is non-distinct from it) include the present.
The -t/tt marked tense, identified as present perfect, corresponds to the following
configuration:

(17) Present Perfect

TP
/ \
UT T’
/ \
T AspP
within / \
RT Asp’
/ \
Asp VP
after / \
ET VP

The Utterance Time is internal to the Reference Time (i.e., the viewpoint), and the Reference
Time is ordered after the Event Time. In other words, a past event is looked at from a present
viewpoint.
 This is how the simple past can be represented in the framework of Demirdache and
Uribe-Etxebarria:

(18) Simple Past

TP
/ \
UT T’
/ \
T AspP
after / \
RTi Asp’
/ \
Asp VP
within / \
ETi VP

The Utterance Time is ordered after the Event Time, which is coindexed with the Reference
Time.
Observe the syntactic representation of the past imperfective:

(19) Past Imperfective

TP
/ \
UT T’
/ \
T AspP
after / \
RT Asp’
/ \
Asp VP
within / \
ET VP

The Utterance Time is ordered after the Reference Time, which is internal to the Event Time.
 Here is Old Hungarian past perfect represented in the framework of Demirdache and
Uribe-Etxebarria:

(20) Past Perfect

TP
/ \
UT T’
/ \
T AspP
after / \
RT Asp’
/ \
Asp VP
after / \
ET VP

The Utterance Time is ordered after the Reference Time, and the Reference Time is ordered
after the Event Time.
According to the proposed analysis, Old Hungarian (the Hungarian language between the
10-16th centuries) possessed a verbal inflection system which included both tense marking
and aspect marking. Aspect had two values: imperfective (with Reference Time within Event
Time), marked by a phonologically null suffix, and perfect (with Reference Time after Event
Time), marked by a -t/tt suffix. Tense also had two values: present (with Utterance Time
within Reference Time), marked by a phonologically null suffix, and past (with Utterance
Time after Reference Time), marked by -a/e.
In the case of the simple tenses (simple present and simple past), the Reference Time
coincides with the Event Time; so I assume no separate AspP projection. As there is only a
tense marker for the verb to pick up, no auxiliary is needed. Observe the morphological make-
up of the simple tenses:

(26)a. Simple present (mond): b. Simple past (monda):

TP TP
/ \ / \
 T VP T VP
0 ‫׀‬ -a ‫׀‬
V V
mond mond

In the case of the complex tenses also marking aspect, the verb combines with the aspect
marker. The tense morpheme, when phonologically non-null, appears on an auxiliary
homophonous with the copula. (The constraint that requires the insertion of lexical support
for the phonologically salient tense suffix is still at work also in present-day Hungarian. As
Bartos (2000:712, 726) formulates it: the verb stem cannot be combined with two analytical
suffixes.) That is:

(27)a. Present perfect (mondott):

TP
/ \
T AspP
0 / \
Asp VP
-tt ‫׀‬
V
mond

b. Past imperfective (mond vala): b. Past perfect (mondott vala):

TP TP
/ \ / \
T AspP T AspP
(val)-a / \ (val)-a / \
Asp VP Asp VP
0 ‫׀‬ -tt ‫׀‬
V V
mond mond

In present-day Hungarian, the inflected verb is subject to the Mirror Principle of Baker
(1985), i.e., the suffixes representing the heads of the morphosyntactic projections extending
the V appear attached to its right-hand side in the opposite order. The universally valid Mirror
Principle was presumably in effect in Old Hungarian, as well. As Bartos (2000) shows,
however, the reverse morpheme order attested on the Hungarian verbal and nominal stems
cannot be the result of successive cyclic head movement to the left, because the occasional
specifiers of the projections extending the head remain on its left-hand side in surface
structure. Therefore, Bartos assumes, instead of head movement, an operation called
‘morphosyntactic merger’, which he defines as follows:

(23) Morphosyntactic Merger

a. A [+affix] category X can morphosyntactically merge with a potential stem Y under
structural adjacency. This yields a word domain {x; y} (where x realizes X, and y
realizes Y) at the level of morphology, and a head-chain <X,Y> in syntax.
b. Structural adjacency
X and Y are structurally adjacent iff
(i) X c-commands Y, and
(ii) there is no Z, such that Z is of the same projection level as X, and X c-commands
Z, and Z c-commands Y.

The structures in (22) are built cyclically, so once e.g. Asp is merged with the VP,
Morphosyntactic Merger creates a word domain {V,Asp}, located under V. The morphemes
of the word are linearized in accordance with the Mirror Principle. Notice that
Morphosyntactic Merger and the Mirror Principle yield the right order in the case of the
complex tenses only if vala is also treated as a suffix, i.e., if the aspect-marked verb plus vala
string is analyzed as a single word. There is also some independent evidence supporting this
assumption. In yes-no questions the interrogative clitic is attached to the finite V, hence it can
be used to test where the right edge of the V is. Present-day Hungarian has preserved a single
relic of the complex verb forms of Old Hungarian: the past conditional, e.g. mondo-tt vol-na
say-PERF.3SG be-COND ‘would have said’. In the case of past conditional predicates, the
interrogative clitic is attached to the whole verbal complex, instead of the V bearing the -t
suffix:

(24)a. Mondtam volna -e nektek a hírt, ha nem igaz?

say-PERF-1SG be-COND-Q you-DAT the news if not true
‘Would I have told you the news if it is not true?’
 b.??Mondtam-e volna nektek a hírt, ha nem igaz?

In the Old Hungarian simple tenses, the agreement morpheme(s) appear on the V plus Tense
complex. In the Subjective Conjugation, the verb only bears a subject agreement suffix. In the
Objective Conjugation, there is also an object agreement morpheme, often fused with the
subject agreement morpheme, marking the presence of a 3rd person definite object:

(25)a. Subjective conjugation

mond-é -k mond-á -nk
say -PAST-1SG say -PAST-1PL
mond-á -l mond-á -tok
say -PAST-2SG say -PAST-2PL
mond-a -0 mond-á -nak
say -PAST-3SG say -PAST-3PL

b. Objective conjugation
mond-á -m mond-á -nk
say -PAST-OBJ.1SG say -PAST-OBJ.1PL
‘I said it’ ‘we said it’
mond-á -d mond-á -tok
say -PAST-OBJ.2SG say -PAST-OBJ.2PL
mond-á -0 mond-á -k
say -PAST-OBJ.3SG say -PAST-OBJ.1PL

Interestingly, in the case of complex tenses, the morphemes representing AgrO and AgrS
appear between the aspect marker and the tense marker, on the aspect-marked V:

(26)a. Subjective Conjugation

mond-t -am val-a mond-t -unk val-a
say -PERF-1SG be-PAST say -PERF-1PL be-PAST
mond-t -ál val-a mond-t - atok val-a
say -PERF-2SG be-PAST say -PERF-2PL be-PAST
mondo-tt -0 val-a mond-t -ak val-a
 say -PERF-3SG be-PAST say -PERF-3PL be-PAST

b. Objective Conjugation
mond-t -a -m val-a mond-t -uk val-a
say -PERF-OBJ-1SG be-PAST say -PERF-OBJ.1PL be-PAST
‘I had said it’ ‘we had said it’
mond-t -a -d val-a mond-t -á -tok val-a
say -PERF-OBJ-2SG be-PAST say -PERF-OBJ-2PL be-PAST
mond-t -a -0 val-a mond-t -á -k val-a
say -PERF-OBJ-3SG be-PAST say -PERF-OBJ-3PL be-PAST

In view of the Mirror Principle, the morpheme order in (26) means that the agreement
projections intervene between AspP and TP, as follows:

(27) TP
/ \
T AgrSP
/ \
AgrS AgrOP
/ \
AgrO AspP
/ \
Asp VP
‫׀‬
V

The problem is that in the case of simple tenses, involving no AspP projection, the agreement
morphemes appear outside the tense morpheme – see (24), i.e., AgrOP and AgrSP presumably
subsume TP. To resolve this contradiction, I tentatively assume that the agreement projections
AgrOP and AgrSP always dominate the first projection extending the verb with an analytic
suffix.

4. The emergence of situation aspect marking

Although the complex tenses of Old Hungarian have survived up till now in the easternmost
Csángó and Székely dialects of Hungarian, they gradually disappeared from standard
Hungarian in the Middle Hungarian period (between 1526-1772). Their loss had been
preceded by a period of neutralization of the functional differences of the various non-present
forms. A few surviving texts also display signs of the loosening of the formal system: in the
Middle Hungarian period we sporadically find past perfect forms like mondo-tt vol-t, in which
not only the verb stem bears the -t perfectivity marker but also the temporal auxiliary bears it
instead of the -a past tense suffix. In literary prose, the various complex verb forms survived
well into the 19th century, but they were only used for variety’s sake, or for an archaic or
elevated effect. Present-day Hungarian has a single non-present tense, marked by -t, the Old
Hungarian perfectivity marker reinterpreted as a past tense suffix.
Parallel with the loss of its complex tense-aspect system, Hungarian has developed
systematic means of marking telicity, the key feature of situation aspect.
As was discussed in Chapter 2, there are three major types of telic predicates in
Hungarian: (i) those expressing a delimited change of state, (ii) those describing a delimited
change of location, and (iii) those expressing creation/coming into being. In the case of type
(i) predicates, telicity is marked by a resultative verbal particle selected by the verb, e.g.:

(28) János le-nyírta/le-nyírja a füvet.

John down cut/down cuts the grass
‘John cut down/will cut down the grass.’

In the case of telic predicates of type (ii), telicity is expressed by a lexically selected
terminative verbal particle, which often doubles a postverbal terminative noun phrase or PP:

(29) János be-ment/be-megy (a munkahelyére).

John in went/in goes his workplace-to
‘John went/will go to his workplace.’

If the verbal particles are omitted, these sentences denote atelic processes:

(30)a. János nyírta/nyírja a füvet.

John cut /cuts the grass
‘John was/is cutting the grass.’
 b. János ment/megy a munkahelyére.
John went/goes his workplace-to
‘John was/is going to his workplace.’

In type (iii) telic sentences, telicity is a consequence of the fact that a verb expressing
creation/coming into being is combined with a non-specific indefinite theme argument (see
chapter 4 of this book). Compare:

(31)a. Sütöttem süteményeket.

baked-I cookies
‘I have baked cookies.’

b. Sütöttem a süteményeket (amikor felhívtál).

baked-I the cookies when phoned-you.me
‘I was baking the cookies (when you phoned me).’

(31a), containing an indefinite object, is telic , whereas (31b), with a definite object, is atelic.
The means of marking telicity, i.e., the category ‘verbal particle’ combinable with any
process verb, on the one hand, and the definite/indefinite articles, on the other hand, are still
missing in 12th-13th century Hungarian texts. In Halotti beszéd és könyörgés ‘Funeral speech
and prayer’, the first surviving longer coherent Hungarian prose, written in 1192-95, none of
the numerous change-of-state and change-of-location predicates has a verbal particle yet. The
telicity of sentences expressing a bounded change is sometimes indicated by the perfective
viewpoint aspect:

(32) es odutta vola neki paradicsumut hazoá. (Forgács et al. 1996b:7)

and give-PERF-OBJ.3SG be-PAST him Paradise-ACC house-for
‘And gave him Paradise for a house.’

When the -a/e-marked simple past is used, the boundedness of the sentence can only be
inferred from the context:

(33)a. es vetevé űt ez munkás világ belé

 and throw-PAST.3SG him this laborious world into
‘and threw him into this laborious world’

b. kit úr ez nopun ez homis világ timnüce beleül mente

whom Lord this day-on this false world’s prison from save-PAST.3SG
‘whom the Lord saved from the prison of this false world on this day’

In present-day Hungarian, we would say, instead of odutta ‘gave’, vetevé ‘threw’, and mente
‘saved’, the particle verbs oda-adta ‘thereto-gave’, be-vetette ‘in-threw’, and ki-mentette ‘out-
saved’.
Sentences expressing a bounded change of state also lack a particle in this 12th century
text:

(34)a. es bulcsássa mend ű bűnét

and forgive-IMPER-OBJ.3SG all his sin
‘and he should forgive all his sins’

b. hugy ougya mend ű bűnét

that solve-IMPER.OBJ.3SG all his sin
‘that he should resolve all his sins’

c. Es szobodochcha űt ürdüng ildetüítűl

and save-IMPER-OBJ.3SG him devil’s harassments-from
‘and he should save him from the devil’s harassments’

Halotti beszéd és könyörgés ’Funeral Speech and Prayer’ contains a single verbal particle in
an interesting function:

(35) Es oz gyimilcsnek úl keseröü vola vizë, hugy turkokat

and that fruit-DAT so bitter was water-its that throat-their-ACC
migé szokosztja vol-a.
PRT burst-IMPERF.OBJ.3SG be-PAST
‘And that fruit had such bitter juice that it was bursting their throats.’
In this example, also quoted under (14), the combination of past imperfective and a telicizing
particle serves to denote the preparatory phase of an achievement situation which fails to
culminate. In such sentences expressing a blocked achievement, both the imperfective
inflection and the verbal particle play a role; the latter denotes the inherent boundedness of
the situation, and the former, the incompleteness of it.
Whereas in Halotti beszéd és könyörgés ‘Funeral Speech and Prayer’ from 1192-95, the
use of the verbal particle is restricted to a very special context, a hundred years later, in
Ómagyar Mária-siralom ‘Old Hungarian Lament of Mary’, delimited motion is already
expressed by a particle plusV complex:

(36) Búabeleül ki -nyuhhad (Forgács et al. (1996a:12))

grief-her-from out pull-IMPER-2SG
‘Pull her out from her grief’

In the Margit legend from 1510, delimited changes of state are also marked by a verbal
particle:

(37)a. az halakat meg-faragja vala

the fishes-ACC PRT carve-IMPERF.OBJ.3SG be-PAST
‘she would carve the fish’

b. az ő kezének bőre meg-hasadoz vala

the her hand’s skin PRT crack.IMPERF.3SG be-PAST
‘the skin of her hands would crack’

Here is an example of a delimited change of location:

(38) el -megyen vala az kapitulumházba

off go-IMPERF.3SG be-PAST the capitulum-house-to
‘she would go to the capitulum house’
(Margit-legenda 1510, in Forgács et al. (1996b:8))

Examples (37)-(38) are imperfective sentences denoting past habits. Their verbal particles
serve to denote the boundedness of the individual actions, whereas the imperfective suffix
expresses habituality. In the case of sentences in the historical past, the verbal particle denotes
the boundedness of the situations.
As the fables of Gábor Pesti from 1536, abounding in perfective verb forms, indicate, the
particle also accompanies perfective verbs:

(39) Az mezőt, hogy ott öttél, mind el -pusztitottad nekem.

the field-ACC that there eat-PERF-2SG all PRT destroy-PERF-OBJ.2SG me
‘By having eaten there, you have destroyed all the field for me.’
(Gábor Pesti, Az farkasról és bárányról ‘On the wolf and the sheep’, in Forgács et
al. (1996b:9))

An interesting „snapshot” of the change taking place in the syntax of late Old Hungarian is
provided by the so-called Müncheni emlék ’Munich relic’, a Hungarian section in a codex
written in the early 16th century presumably by a German monk interested in languages (for
details, see Haader (2004)). The relic contains, among others, two versions of the prayers
Pater noster and Ave Maria in Hungarian. The scribe first transscribed phonetically what he
heard from a Hungarian person reciting the two prayers, and then he copied the prayers from a
written Hungrian source. Whereas the written prayers, representing an earlier stage of Old
Hungarian, contain practically no verbal particles, in the oral versions, representing the actual
usage of the early 16th century, the verbs bocsát ‘forgive’ and szabadít ‘liberate’ already
occur with the verbal particle meg.
Since the 14th century, more and more accomplishment and achievement verbs have come
to be lexicalized with a verbal particle, until we have reached the present stage when
practically all verbs denoting a delimited change of state or location take a particle (at least in
the spoken language). In the course of this development, simple particleless verbs denoting a
change of state or change of location have assumed an unambiguously atelic reading.
Verbs of creation and coming into being assume a telic interpretation if their theme
argument, whose creation or coming into being is asserted, is a non-presupposed, non-specific
indefinite noun phrase. In the following pair of examples representing present-day Hungarian,
the (a) sentence is unambiguously atelic, and the (b) sentence is unambiguously telic:

(40)a. Jöttek vendégek.

came guests
‘Guests have come/came.’
 b. (Éppen) Jöttek a vendégek (amikor telefonáltál).
(just) came the guests (when phoned-you)
‘Guests were coming (just when you phoned).’

The system of determiners that makes this type of aspectual distinction possible emerged in
the late Old Hungarian period. The definite article, derived from the demonstrative pronoun,
appeared in the 14th century; and the indefinite article, derived from the numeral one,
appeared in the 15th century. In the München relic from the early 16th century, the
transcription of the oral Pater noster already says szabadíts meg minket a gonosztúl ‘save
PRT us the evil-from’, whereas in the more conservative written version gonosztúl ‘evil-
from’ has no article yet. The slow extension of the articles to their various present-day
functions reached a stage resembling the usage of present-day Hungarian in the 16th century
(cf. I. Gallasy (1991, 1992)). Thus at the beginning of the period of Middle Hungarian, both
means of marking telicity, the verbal particle and the system of definite/indefinite
determiners, were ready in the language.
The emergence of the system of telicity marking was soon followed by the attrition of the
system of complex tenses. Middle Hungarian was the period of the gradual loss of the
functional differences between the complex verb forms marking both tense and viewpoint
aspect. By the 19th century, the present perfect (e.g. mondo-tt) had been reinterpreted as a
past tense, and the past tense forms, i.e., the simple past monda, the past imperfective mond
vala, and eventually even the past perfective mondott vala came to be used as stylistic
variants of the regular past tense marked by -t. That is, Hungarian, representing a language
type with a grammaticalized viewpoint aspect in the period of Old Hungarian, went through a
typological shift in the period of Middle Hungarian, as a result of which Modern Hungarian
has become a language with a grammaticalized situation aspect.
The syntactic structure of the Modern Hungarian sentence involves a PredP projection
according to the evidence presented in Chapter 2. PredP dominates VP, with the V raised to
Pred. Spec,PredP is filled alternatively by a verbal particle or a bare nominal complement.
The PredP projection actually must have existed in the Old Hungarian period, as well.
Although Old Hungarian – especially its earlier stages – used the verbal particle only
sporadically, the use of non-referential, predicative bare nominals was wide-spread. They
appeared preverbally, presumably in the same Spec, PredP where they appear in present-day
Hungarian. For example:
(41)a. isa por es homu vogymuk
indeed dust and ash are-we
‘indeed we are dust and ashes’
(Halotti beszéd és könyörgés ‘Funeral speech an prayer’ 1192-1995, in Forgács et
al. (1996b:7))

b. hugy szűz leány fiot szülhessen

that virgin girl son-ACC bear-POT-SUBJUNC-3SG
‘that a virgin girl should be able to bear a son’
(Königsberg fragment 1350, quoted by R. Hutás (1991:666))

As argued in Chapter 9, Spec,PredP is also the position of identificational focus. In Old

Hungarian, just as today, identificational foci, e.g. WH-phrases, stood immediately
preverbally, presumably in Spec,PredP:

(42)a. Látjátuk feleim szümtükhel, mik vogymuk?

see-you fellows-my eyes-your-with what are-we
‘Do you see, my fellows, with your eyes what we are?’
(Funeral speech and prayer, 1192-1195, in Forgács et al. (1966b:7))

b. Menyi milosztben terümtevé elevé miü isemüköt, Ádámut

how.much grace-in created earlier our ancestor-ACC, Adam-ACC
‘In how much grace he created our ancestor, Adam’
(Funeral speech and prayer, 1192-1195, in Forgács et al. 1966b:7))

c. Kik vattok ti?

who are you
(Jókai-kódex 1448:30, quoted in Pólya (1991:592))

According to the evidence of the examples in (41) and (42), Spec,PredP, the position
harboring the verbal particle in present-day Hungarian, was already available when the verbal
particle became widely used..
 When Spec,PredP came to be associated with telicity marking (and indirectly with
viewpoint aspect, as well, as will be argued in Section 5 below), the projection assumed a
kind of aspectual function; Asp may have become part of the feature specification of Pred.
This paved the way for the reinterpretation of the original AspP projection as a TP, and for the
disappearence of the original TP projection.
A similar process must have taken place in various Slavic languages (except Bulgarian,
which has both developed particle-like verbal prefixes, and has preserved its complex tenses –
see Bertinetto (2001)). German may very well be a language undergoing this kind of type-
shift at present. On the one hand, more and more German telic verbs are associated with a
telicizing particle. On the other hand, the functional differences between the perfect and the
imperfect past tenses have been neutralized, and speakers use the different non-present forms
as stylistic variants in most dialects.
The fact that the appearance of situation aspect marking was followed by the
disappearance of viewpoint aspect marking both in Hungarian and in several other languages
suggests that these two processes are not independent of each other. Perhaps the overt
marking of (a)telicity makes the systematic morphological marking of (im)perfectivity
redundant; perhaps viewpoint aspect can also be inferred from situation aspect. This is the
question that section 5 will address.

5. Can viewpoint aspect be inferred from situation aspect?

Although it is a basic tenet of the two-component aspect theory of Smith (1991) that situation
aspect and viewpoint aspect are independent of each other, in fact most situation types imply
a particular viewpoint. For example, true accomplishments and achievements can only be
described from the perfective viewpoint. Both Erzsi meg-főzi a vacsorát ‘Elisa PRT cooks the
supper’ and Erzsi meg-főzte a vacsorát ‘Elisa PRT cooked the supper’ are perfective. What
distinguishes an accomplishment from a process is the inherent endpoint of the former. By
excluding the final, defining element of an accomplishment situation, the imperfective
viewpoint – extending over an internal segment of the situation – neutralizes the difference
between an accomplishment and a process. Thus the omission of the verbal particle meg
yields the process sentence Erzsi főzte a vacsorát ‘Elisa was cooking the supper’. In the case
of achievements, the inherent endpoint is simultaneous with the initial point of the process,
hence the omission of the endpoint, i.e., an imperfective viewpoint, is impossible. Compare:

(43)a. Meg találtam a helyes választ.

 PRT found-I the correct answer
‘I found the correct answer.’

b.*(Éppen) találtam a helyes választ (amikor az óra véget ért).

(just) found-I the correct answer (when the class ended)
‘I was (just) finding the correct answer (when you phoned me).’

States, on the other hand, are necessarily imperfective. A state situation is claimed by Smith
(1991) not to include the initial point and the endpoint of the given state – hence it follows
that it cannot be represented from a perfective point of view. Indeed, both János tudja a
matematikát ‘John knows (the) mathematics’ and János tudta a matematikát ‘John knew (the)
mathematics’ are imperfective,
In the case of processes, the speaker can include or exclude the initial point and the
endpoint of the situation in his/her viewpoint at will. In the case of processes involving a
specific object, like those in (44), the fact that no verbal particle is spelled out indicates that
the situation is unclosed, hence imperfective.

(44)a. János írja a dolgozatot.

John writes the paper
‘John is writing the paper.’

b. János írta a dolgozatot.

John wrote the paper
‘John was writing the paper.’

In the case of unergatives predicates, like those in (45), the possibility of a verbal particle is
excluded; it is the bare verb that can be interpreted either perfectively or imperfectively.1 The
present tense form makes the imperfective reading likelier (45a), whereas the past tense form
favors the closed, perfective interpretation (45b) – although an appropriate time adverbial can
also elicit the perfective reading of the present tense form (45c), and the imperfective reading
of the past tense form (45d):

(45)a. János teniszezik.

John plays.tennis
 ‘John is playing tennis.’

b. János teniszezett.
John played.tennis
‘John played tennis.’

c. János délután teniszezik, aztán moziba megy.

John afternoon plays.tennis then cinema-to goes
‘In the afternoon, John plays tennis, then he goes to the cinema.’

d. János éppen teniszezett, amikor kerestem.

John just played.tennis when searched-I-him
‘John was playing tennis when I looked for him.’

In sum, the major types of situations correspond to the following viewpoints, at least in the
unmarked cases:

(46) accomplishment –> perfective

achievement –> perfective
state –> imperfective
process –> imperfective or perfective

The question is if the correspondences under (46) represent the unmarked cases or these are
the only the possibilities. In the literature, there are also telic predicates represented from an
imperfective viewpoint and states represented from a perfective viewpoint. I will claim that
these cases, too, can be reduced to the correspondences in (46), because in such sentences the
predicate undergoes a situation aspect shift prior to viewpoint aspect interpretation. That is, in
an apparently telic imperfective sentence, the lexically determined telicity of the predicate is
changed to atelicity before the imperfective viewpoint can be imposed on it. Similarly, an
apparently perfective state is not a state any more; it has been telicized. Thus an unexpected
viewpoint aspect is always a consequence of a shift of the situation type.
The most common situation type shifts are the following:

i. From state to accomplishment/achievement

Most stative predicates, e.g. van ‘be’, tud ‘know’, szeret ‘love, like’, lát ‘see’, tud ‘know’ can
be transformed by means of a resultative verbal particle into telic change-of-state predicates,
for example: meg-van, meg-szeret, meg-lát, meg-tud. As is demonstrated in detail in chapter
XX, these predicates represent the state that they denote as a resultant state, attained as a
consequence of a change. Whereas stative predicates are imperfective, their telic counterparts
are naturally perfective. Compare:

(47)a. Volt pénz az építkezéshez.

was money the building.project-for
‘There was money for the building project.’

b. Két héten belül meg-volt a pénz az építkezéshez.2

two week within PRT was the money the building.project-for
‘The money was obtained for the building project within two weeks.’

(48)a. János haláláig szerette Évát.

John death-his-until loved Eve
‘John loved Eve until his death.’

b. János rövid idő alatt meg-szerette Évát.

John short time in PRT loved Eve
‘John came to love Eve in a short time.’

(49)a. Gábor látja a várat az ablakából.

Gabriel sees the castle the window-his-from
‘Gabriel sees the castle from his window.’

b. Gábor hirtelen meg-látta a barátját az ablakából.

Gabriel suddenly PRT saw his friend the window-his-from
‘Gabriel suddenly caught sight of his friend from his window.’

(50)a. Sándor tudja az eredményt.

Alexander knows the result
‘Alexander knows the result.’
 b. Sándor tíz perc alatt meg-tudja az eredményt.
Alexander ten minute in PRT knows the result
‘Alexanderlearns the result in 10 minutes.’

ii. From accomplishment to process

Accomplishment predicates are complex predicates consisting of a process component and a
resulting state component. The removal of the resulting state yields an imperfective process
predicate.3 If the state component is represented by a resultative verbal particle that does not
alter the lexical meaning of the verb apart from telicizing it, the removal of the resulting state
is realized by the deletion of the verbal particle:

(51)a. István fél óra alatt el -olvasta az újságot.

Stephen half hour in PRT read the newspaper
‘Stephen read the newspaper in half an hour.’

b. István fél óra hosszat olvasta az újságot.

Stephen half hour long read the newspaper
‘Stephen was reading the newspaper for half an hour.’

If the verbal particle also contributes to the lexical meaning of the predicate, its ‘removal’
means its realization in a postverbal complement position. Compare:

(52)a. Géza tíz perc alatt fel-ment a toronyba.

Geza ten minute in up went the tower-to
‘John went up to the tower in ten minutes.’

b. Géza éppen ment fel a toronyba, amikor meg-láttam.

Geza just went up the tower-to when PRT saw-I-him
‘John was just going up to the tower when I caught sight of him.’

(53)a. Rozi fél óra alatt össze -írta a neveket.

Rosy half hour in together wrote the names
‘Rosy wrote down the names in half an hour.’
 b. Rozi egész délután írta össze a neveket.
Rosy whole afternoon wrote together the names
‘Rosy was writing down the names the whole afternoon.’

A postverbal particle can be supplemented with the morpheme -fele ‘-wards’, which
apparently serves to indicate that the given element is a directional adverb instead of a
telicizing particle. That is:

(54)a. Géza éppen ment felfele a toronyba, amikor megláttam.

‘Geza was just going upwards in the tower when I caught sight of him.’

b. Rozi egész délután írta összefele a neveket.

‘Rosy was writing downwards the names the whole afternoon.’

The imperfective aspect of the sentences in (51b), (52b), (53b), and (54) is obviously a
consequence of the type shift of the accomplishments to a process.
In the case of achievements, such a type shift is not possible. Since in their case the
process and the resultant state are essentially simultaneous, the removal of the resultant state
does not leave a situation interpretable as a process:

(55)a.*János éppen érte el a hegycsúcsot, amikor ki-tört a vihar.

John just reached PRT the mountaintop when out broke the storm
‘John was just reaching the mountaintop when the storm broke out.’

b.*Fleming éppen találta fel a penicillint, amikor …

Fleming just invented PRT the penicillin when…
‘John was just inventing penicillin when…’

The following predicates do contain an independent, albeit short, process component

separable from the resultant state, so they are accomplishments rather than achievements; that
is why the particle potentially denoting the resultant state can be turned into a directional
complement:
(56)a. János éppen csukta be(fele) az ajtót, amikor…
John just closed in(wards) the door when…
‘John was just closing the door, when…’

b. Miklós éppen kapcsolta be(fele) a tévét, amikor…

Nicholas just turned in(wards) the TV when…
‘Nicholas was just turning on the TV when…’

iii. From accomplishment/achievement to state

The following constructions assert a state which is the resultant state of a previous
accomplishment or achievement.

(57)a. A pad be van festve.

the bench in is paint-ADV
‘The bench is painted/is in a painted state.’

b. Az ajtó be van csukva.

the door in is close-ADV
‘The door is closed/is in a closed state.’

c. A levél meg van írva.

the letter PRT is write-ADV
‘The letter is written/is in a written state.’

I assume that this construction is a raising construction, with the copula functioning as the
raising verb, that is, e.g. (57a) is derived from the following structure:

(58) van [a pad be-festve]

Although these sentences contain an accomplishment or achievement predicate, it is buried

under an AdvP. The head of AdvP is the suffix -va/ve, which derives an adverbial participle
phrase from the PredP, and is realized on the verbal head:

(59) van [AdvP a pad [AdvP -ve [PredP be [fest]]]

Eventually, not only the subject of AdvP undergoes Raising, but also the verbal particle is
moved into the matrix Spec,PredP - see (60a), unless the matrix verb is preceded by a focus or
a negative particle – see (60b):

(60)a. A padi bej van [AdvP ti tj festve]

b. A padi CSÚNYÁN van [AdvP ti be -festve]

the bench uglily is PRT painted
‘The bench is painted in an ugly way.’

(As has been made clear by Szendrői (2003), van, a light verb, cannot bear phrasal stress,
therefore it cannot occupy the leftmost position of a predicate phrase, the locus of main stress,
but must be preceded by a verbal particle, a focus, or a negative particle.)
I assume that the adverbial suffix in the construction in (57)-(60) has a detelicizing role,
turning the accomplishment denoted by the PredP into a (resultant) state. Since the matrix
predicate is also stative, it does not change the atelic nature of AdvP. Therefore, the viewpoint
aspect of such sentences is imperfective:

(61) Az ablak fél óra hosszat ki volt nyitva.

the window half hour long out was opened
‘The window was opened/was in an opened state half an hour long.’

In Hungarian, the copula has an imperfective variant (van), and a perfective variant (lesz), as
well. This seems to be a relic of the earlier viewpoint-aspect marking period of the language.
(The present tense form of the perfective copula, lesz, also functions as the future tense form
of the imperfective copula.) If the matrix predicate of e.g. (57c) is replaced by lesz, then,
naturally, the viewpoint aspect of the sentence becomes perfective:

(62) A levél két perc alatt meg lett írva.3

the letter two minute in PRT be.PERF-PAST write-ADV
‘The letter has been/was written in two minutes.’

iv. From accomplishment/achievement to process

The predicates of the following sentences are telic achievement/accomplishment predicates.
Nevertheless, the aspect of the sentences is imperfective, as is indicated by the durative
adverbials:

(63)a. János két hete haza-kiséri Marit iskola után.

John two week-for home accompanies Mary school after
‘For two weeks, John has been seeing Mary home after school.’

b. Éva férje halála óta minden nap ki -megy a temetőbe.

Eve her.husband’s death since every day out goes the cemetery-to
‘Since her husband’s death, Eve has been going to the cemetery every day.’

In these sentences, the telic predicate denotes a single delimited event over which
quantification is performed. (63a) contains an invisible habitual/generic operator, and (63b)
contains a universal quantifier. I assume that these operators create complex atelic situations,
processes or states, which are subject to the generalization in (46), allowing or requiring an
imperfective reading.

5. Conclusion
This chapter has argued that in the period of Middle Hungarian, the Hungarian language has
gone through a typological shift: it has developed a system of marking telicity, and parallel
with that process, it has lost its complex tense-aspect system marking viewpoint aspect. It has
been demonstrated that the lack of overt viewpoint aspect marking in present-day Hungarian
does not result in the loss of any aspectual information. In the unmarked case, predicates
marked as telic have the perfective viewpoint, whereas atelic predicates have the imperfective
viewpoint. There is a single situation type: the process, in the case of which viewpoint aspect
cannot be directly inferred from situation aspect – given that a process situation is equally
likely to be represented from the imperfective and the perfective viewpoints. Apparently,
unergatives, which cannot be telicized, and transitive and unaccusative verbs behave
differently in this respect. As for unergatives, a present tense form is understood as
imperfective, and a past tense form is understood as perfective in the unmarked case. In the
case of transitive and unaccusative predicates, which have the ability to take a telicizing
verbal particle, the imperfective reading is the unmarked choice in both tenses.
 Perfective atelic situations and imperfective telic situations represent special cases, which
arise as a consequence of a change in the temporal structure, and hence, the (a)telic feature of
the situation. That is, perfective atelics are, in fact, atelic situations turned telic, hence
perfective. Similarly, imperfective telic situations are imperfective because the originally telic
situation has been turned atelic. These changes are marked by syntactic or morphosyntactic
means, therefore the modified aspectual value of the sentence is inferable. Thus perfective
states are not simple states; they are resultant states arising as a consequence of a previous
change, which is indicated by the stative verb being associated with a resultative particle.
Imperfective telic predicates can be of at least three different kinds. The removal – or the
postposition into complement position – of the resultative element of an accomplishment
predicate yields an imperfective process predicate. An accomplishment or achievement
predicate can be transformed into a state by the addition of an adverbial participle suffix, and
by the subordination of the resulting participle phrase to a stative raising verb. Quantification
over an accomplishment or achievement situation e.g. by means of a universal quantifier can
also result in an open, atelic complex situation requiring an imperfective interpretation.
The possibility of deriving viewpoint aspect from situation aspect casts doubt on claims
that situation aspect and viewpoint aspect represent two independent systems.4

References
Baker, Mark 1985. ‘The Mirror Principle and morphosyntactic explanation’. Linguistic
Inquiry 16, 373-415.
Bánhidi, Zoltán 1941. A magyar nyelv összetett igealakjai. A Magyar Nyelvtudományi
Társaság Kiadványai 63. Budapest.
Bartos, Huba 2000. ‘Az inflexiós jelenségek szintaktikai háttere’. In Kiefer Ferenc (ed.),
Strukturális magyar nyelvtan 3. Morfológia. 653-761. Budapest: Akadémiai Kiadó.
Benkő, Loránd (ed.), 1991. A magyar nyelv történeti nyelvtana I. A korai ómagyar kor és
előzményei. Budapest: Akadémiai Kiadó.
Benkő, Loránd (ed.), 1992. A magyar nyelv történeti nyelvtana II/1. A kései ómagyar kor.
Morfematika. Budapest: Akadémiai Kiadó.
Bertinetto, Pier Marco 2001. ‘On a frequent misunderstanding in the temporal-aspectual
domain: the perfective – telic confusion’. In Carlo Cecchetto, Gennaro Chierchia, Maria
Teresa Guasti (eds.), Semantic interfaces. 177-209. Stanford: CSLI.
Bohnmeyer, Jürgen and Mary Swift in press. ‘Default aspect: The semantic interaction of
aspectual
viewpoint and telicity’. To appear in A. van Hout, H. de Swart, and H. Verkuyl (eds.),
Perspectives on Aspect. Utrecht: UiLOTS Working Papers.
Demirdache, Hamida and Myriam Uribe-Etxebarria 2004. ‘The syntax of time adverbs’. In
Jacqueline Guéron and Jacqueline Lecarme (eds.), The Syntax of Time. 143-180.
Cambridge, Mass.: MIT Press.
E. Abaffy, Erzsébet 1992. ‘Az igemód- és igeidőrendszer’. In Loránd Benkő (ed.), 1992. 120-
183.
Forgács, Anna et al. (eds.) 1996a. Irodalmi szöveggyűjtemény érettségizőknek I. A magyar
költészet évszázadaiból. Budapest: Helikon.
Forgács, Anna et al. (eds.) 1996b. Irodalmi szöveggyűjtemény érettségizőknek II. A magyar
szépprózából és verses nagyepikából. Budapest: Helikon.
Haader, Lea 2004. Nyelvi szempontú megjegyzések a Müncheni emlékhez [Linguistic
remarks on the München relic], ms, to appear in Magyar Nyelv.
Hornstein, Norbert 1993. As Time Goes By. Tense and Universal Grammar. Cambridge,
Mass.: MIT Press,
I. Gallasy, Magdolna 1991. ‘A névelő és névelő-előzmény’. In Loránd Benkő (ed.), 1991.
461-475.
I. Gallasy, Magdolna 1992. ‘A névelők’. In Loránd Benkő (ed.), 1992. 716-770.
Piñon, Christopher 1995. ‘Around the progressive in Hungarian’. In István Kenesei (ed.),
Approaches to Hungarian 5, 153-192.
Pólya, Katalin 1991. ‘A mondat és fajai.’ In Loránd Benkő (ed.), 1991. 587-592.
Reichenbach, Hans 1947. Elements of Symbolic Logic. Reprinted in 1966 by Free Press,
New York.
R. Hutás, Magdolna 1991. ‘Az igei szerkezetek’. In Loránd Benkő (ed.), 1991. 649-472.
Smith, Carlota 1991. The Parameter of Aspect. Dordrecht: Kluwer.
Szendrői, Kriszta 2003. ‘A stress-based approach to the syntax of Hungarian focus’. The
Linguistic Review 20, 37-78.

Notes
1 Recall from chapter 2 that the verbal particle is a secondary predicate predicated of the
theme argument; hence it is not licensed in the case of verbs taking no theme.
 2 The verb meg-van ‘PRT-is’ is, in fact, ambiguous. It can also be used statively, to denote the
existence of a specific subject, e.g.:

(i) Jánosnak meg-van az Anna Karenina.

John-DAT PRT is the Anna Karenina
’John has Anna Karenina.’

3 A similar proposal was put forward by Piñon (1995). In the derivation of the progressive
aspect proposed by him, the first step is performed by a PR operator which creates process
predicates from event predicates by separating the process stage of events.

4 The construction in question resembles the passive, yet it is more restricted than e.g. the
English passive. Its aspect is fixed (but see the discussion of (62)), and only verbs denoting a
delimited change-of-state can occur in it.

5 A similar point is made by Bohnmeyer and Swift (in press).

View publication stats