**Dov Yarden Hebrew and Arabic linguistic Seminar at the École Normale**
Workshops on Semitic Syntax and Morphology
# Biblical Hebrew noun phrases as adverbials
Biblical Hebrew (BH), like English, employs noun phrases (NPs) in
certain kinds of adverbials (e.g. I will arrive \[next week\]).
However, it appears that BH allows a wider range of adverbial NPs than
English does even though most adverbials in BH are prepositional
phrases. And yet, if an NP is neither the subject nor an object of the
verb, is it necessarily an adverbial? The basis for this paper is a
comprehensive study of Genesis through Deuteronomy which has yielded
several hundred examples. Until now the NPs catalogued in this study
have tended to be merely listed in taxonomies of ‘adverbial
accusatives.’ But such treatment obscures the fact that some of these
NPs have quite different syntactic functions. This paper is meant to
offer some first steps toward determining which NPs are adverbial and
how their syntax differs from those that are not. ‘Adverbial’ NPs
should first of all be distinguished from arguments and secondary
predicates. It also turns out that ‘modifier’ is a better term for
what remains. What have traditionally been called adverbials can be
broken down into event-external modifiers, event-internal modifiers,
and frame-setting modifiers by assuming an event semantics approach.
The benefit of this subdivision is that it allows straightforward
analysis of some of the more puzzling ‘accusatives.’
1. > [Introduction 1](#introduction)
2. > [Theoretical Assumptions 2](#theoretical-assumptions)
3. > [Bare NP Adverbials 2](#bare-np-adverbials)
1. > [Terminology and Traditional Expectations
2. > [English Adverbial NPs 3](#english-adverbial-nps)
3. > [BH Adverbial NPs: A First Pass
4. > [Ruling Out Arguments and Secondary Predicates
4. > [NP Arguments: A Valency Approach
5. > [NP Secondary Predicates 6](#np-secondary-predicates)
5. > [Three Types of Locative ‘Adverbial’ NP
6. > [An Event Semantics Approach to Modifiers
7. > [The Syntax and Semantics of Locative Modifiers
8. > [Locative NP Modifiers in BH 8](#locative-np-modifiers-in-bh)
6. > [Conclusion 9](#conclusion)
*This research was supported by the Social Sciences and Humanities
Research Council of Canada.*
There is a long tradition that treats Biblical Hebrew noun phrases
according to the three Semitic cases: nominative, genitive, and
accusative. Within this tradition the accusative case is considered to
be a broad category that subsumes every kind of phrase that is in some
way ‘subordinate’ to the verb. As shown in Table 1 of the handout, most
grammars divide the accusative case into two major types. Although they
vary in their terminology, all recognize that the first category of
accusatives is somehow part of the core of the clause, while the second
category is somehow peripheral.
Table 1: Basic Categorization of ‘Accusatives’
#### Scholar(s) Category 1 Category 2 Category 3
Gesenius- Kautzsch-Cowley (1910)
Direct Subordination of the Accusative
Looser Subordination of the Accusative
Joüon (1947)1 Direct Accusative Indirect Accusative
Objective Accusative & Double Accusative2
Arnold-Choi (2003) Object Accusative Adverbial Accusative
Van der Merwe- Naudé-Kroeze (2017)
Boulet Arguments Modifiers Secondary
This paper summarizes the direction of my dissertation, a project that
began with the suspicion that the standard grammars leave something to
be desired when it comes to ‘accusatives.’ In particular, I focus on the
second, ‘peripheral’ category from Table 1 in order to determine what
properly belongs there. A systematic approach to this question is shown
to be necessary by observing the extent to which the grammars differ on
the analysis of individual cases. They disagree on which
1. See also the revision, Joüon-Muraoka (2006), which maintains the
2. Waltke-O’Connor actually list objective and double Accusatives
separately, but without good reason in my opinion since double
accusatives are associated with the ‘complex transitivity’ of
certain verbs (§10.2.3a), just as single objects are associated with
verbs according to the verb’s features (§10.2.1a).
examples belong under category 1 and which under category 2. They also
disagree widely on how many subcategories there should be and what they
should be called. Although the title of the paper uses the term
‘adverbial noun phrases,’ a term familiar from works like
Waltke-O’Connor for category 2, I will suggest that this term needs to
be reevaluated. There are two main points to this paper. First, close
inspection of the candidates for category 2 reveals that some noun
phrases belong neither in category 1 nor in category 2. This means that
we need at least three categories and not two. I label these categories:
arguments, modifiers, and secondary predicates. The second point is that
what are traditionally called adverbials are better called modifiers.
‘Modifier’ is a more general term, while ‘adverbials’ are
considered to be a subtype of modifier.
# Theoretical Assumptions
Before rushing into the matter at hand, it is well to introduce the
theoretical assump- tions behind this study. The linguistic framework
adopted here is the generative Minimalist Program (e.g. Chomsky 1993,
1995, 2000) as realized in conjunction with Distributed Morphology (e.g.
Halle and Marantz 1993, 1994; Harley 2012). My take on argument
structure combines a valency approach to BH verbs3 with the syntactic
structures proposed by Cuervo (2003, 2015), with modification of the
latter to agree with the unified approach to predication proposed by
Bowers (1993, 2001).
3. # Bare NP Adverbials
1. ## Terminology and Traditional Expectations
An ‘adverbial,’ as typically understood, is a word or phrase which
modifies the verbal action in some way (hence ad + verb(ial)). In
general, adverbials may belong to any category, whether noun phrases,
prepositional phrases, or complementizer phrases. The label ‘adverbial’
should be understood as a descriptor of a phrase’s semantic function.
If one is inclined rather to describe the syntax of an adverbial, one
may refer to it as an ‘adjunct,’ reflecting the fact that adverbials are
thought to be ‘adjoined’ to other phrases. Fundamental to the label
‘adjunct’ is the idea that the adjoined phrase is grammatically
optional. The presence or absence of an adjunct does not affect the
grammaticality of a clause in the same way that the omission of an
‘argument’ does. Arguments are regarded here as obligatory
participants in a valency pattern of a specific verb.
3\. For the valency approach applied to BH see Malessa (2006), Cook
(2012a, 2012b, 2014), and Wilson (2014).
Neither arguments nor adjuncts are constrained to any one lexical
category. Nev- ertheless, there is a general expectation that noun
phrases ought to be arguments (and not adjuncts) and that arguments are
prototypically noun phrases.
## English Adverbial NPs
When considering adverbial noun phrases in Biblical Hebrew, it is
instructive to compare potential cases with the stock of adverbial noun
phrases in a more familiar language like English. Larson (1985) has
found that in English, adverbial noun phrases may belong to one of four
basic categories: time, location, direction, and manner. Not only are
the semantic types severely limited in English, but even within
categories, Larson found that nouns used adverbially in English are
restricted to a limited set of lexical items. The following examples
show that some bare noun phrases are grammatical as adverbials in
English, while others require the use of a preposition.
We may say ‘John will arrive \[sometime next week\],’ but not ‘John will
arrive \[that occasion\].’ We must say ‘John will arrive
\[on that occasion\].’ Similarly, we may
say ‘You have lived \[someplace warm and sunny\],’ but not ’You have
lived \[Germany\].’ Or again, we may say ‘We are headed \[that
direction\],’ but not ‘We are headed \[that course\].’ Larson found that
manner adverbial noun phrases in English are the most limited, being
restricted to phrases with the head noun way.
Without assuming that Biblical Hebrew must conform to the same
restrictions as English, the English data does give us interesting
questions to consider. First, should we expect adverbial noun phrases in
Biblical Hebrew to be restricted to certain semantic categories like
time, location, direction and manner? Second, is it the case that
Biblical Hebrew adverbial noun phrases are limited to a specific set of
lexical items? To these questions I add this one: is it the case that a
noun phrase with prototypically ‘adverbial’ semantics (say, those
belonging to Larson’s four categories) is necessarily an adverbial? In
what follows I argue that one needs to pay attention to the semantics,
not only of the noun phrase under consideration, but also of the verb.
It will be shown, for example, that some locative noun phrases are not
adverbial at all, but rather the arguments of a verb that selects a
location as argument.
## BH Adverbial NPs: A First Pass
As a first pass, let us consider some examples identified as ‘adverbial’
by Waltke- O’Connor which fit well with Larson’s categories,
specifically the accusatives of time, manner, and place listed in the
upper portion of Table 2 on the second page of the handout. The noun
phrase in question is underlined for clarity.
The first example is from Jeremiah 28:16, where the noun phrase
the point in time of the hearing event. It reads as follows:
?C֣ה ֵמת ֽ?י־ס ָ ֥רה ד ַ ֖fi ְר? אל־יה ָוה (1)
ַה ָ% ָנ ה֙
‘You (are going to) die this year because you have spoken rebellion
against the lord.’ (Jer 28:16) (Waltke-O’Connor 1990, §10.2.2c)
Exodus 20:9 attests another temporal modifier, this time with the noun
?֤ specifying temporal duration.
(2) ֮ ֹבד
‘Six days you may work.’ (Exod 20:9) (Waltke-O’Connor 1990, §10.2.2c)
Temporal modifiers are the most common and also the most obvious type,
ac- counting for approximately half of the adverbial noun phrases in the
Pentateuch, and the grammars tend to agree on their classification. In
general, these either specify the point in time of the action or the
duration of an action as already seen in the examples.
Biblical Hebrew also allows noun phrase manner adverbials. A typical
example could be called the ‘as one’ construction which refers to
multiple people acting in concert and without divisions, as though a
single person . The embedded infinitival clause from Zephaniah 3:9 is
such an example.
Lit: ‘...(so that they) serve Him (with) one shoulder.’ (Zeph 3:9)
(Waltke- O’Connor 1990, §10.2.2e)
The noun phrase ֽחד ָ
ֶ ?, ‘one shoulder,’ modifies the action of
service. It is a man-
ner adverbial because it answers the question ‘how?’. In context, God is
promising to unify the peoples to worship and serve Him as one.
Spatial noun phrases are also common. These may include the location of
the action, the distance covered, and the direction of movement. Unlike
in English, there is as yet no sign that there is a restriction on noun
phrases with spatial semantics that may occur as modifiers. Although in
1 Kings 8:32 we might expect a prepositional
ואתה, here, as also attested elsewhere, a
phrase will do for the location of the hearing event.
ַה ַמ֣ע ??
‘And you, listen (in) heaven.’ (1 Kgs 8:32) (Waltke-O’Connor 1990,
The foregoing examples are meant to introduce the concept of noun phrase
mod- ifiers on relatively solid ground. It may not be a coincidence that
these correspond to Larson’s categories for English. After sifting
through other candidates, this paper proposes that Biblical Hebrew
allows modifiers of three types: time, manner, and place.
# Ruling Out Arguments and Secondary Predi- cates
Many of the noun phrases analysed by the grammars as belonging to
– the traditional ‘adverbial’ or the syntactic ‘adjunct’ – actually
belong to category one: these are what I call ‘arguments.’ Others are
neither ‘arguments’ nor ‘modifiers,’ but rather ‘secondary predicates,’
the third category I am proposing in this paper.
## NP Arguments: A Valency Approach
Arguments are the more familiar category, so I begin here. Whereas the
traditional grammars list categories of accusative based largely on the
semantics of the noun phrase in question, a better approach also
considers the semantics of the verb, since the grammaticality of a
clause depends in part on the compatibility of the arguments with the
verb. The approach I have in mind is called a valency approach, such as
the one advocated by John Cook. Valency refers to the ‘combining
capacity’ of a verbal lexeme. According to their basic meaning, verbs
combine with arguments in particular patterns. Each pattern for the verb
selects a specific number of arguments of specific types.
In general, valency patterns combine with as few as zero and as many as
three ar- guments and are said to be avalent, monovalent, bivalent, or
trivalent corresponding to the number of arguments. Some verbs are
associated with a single valency pat- tern, while others may occur in
multiple distinct combinations. In Biblical Hebrew, valency applies to
the combination of a root and a binyan, or ‘stem.’
A better understanding of a verb’s valency can allow us to identify
which phrases are (selected) arguments and which are (optional)
adjuncts. In cases where a verb is infrequently attested it may be
possible to use close synonyms, to the extent that there are any, for
Let us start with a common verb, the Hebrew
ב?י, meaning ‘to sit or
stay.’ By knowing that the Qal of
ב?י has a common bivalent
pattern meaning ‘X stayed (at place) Y,’ we can avoid the mistake made
by Waltke-O’Connor when they list
תח־ה ֖ ֹאהל
from Genesis 18:1 as an adverbial accusative of place (§10.2.2b).
ה ֹחם ?
‘And he (was) sitting (at) the entrance of the tent during the hot
part of the day.’ (Gen 18:1) (Waltke-O’Connor 1990, §10.2.2b)
The bivalent pattern in question selects a Doer and a Location. The
Location may be either a noun phrase or a prepositional phrase. Either
way, it is called for
and licensed by the verb. Therefore, הל
תח־ה ֖ ֹא
is an argument, and this despite
the fact that the noun phrase has locative semantics. It takes
understanding the verb’s semantics to properly categorize it.
The traditional approach to noun phrases causes a proliferation of
category terms. The following two examples have been dubbed adverbial
accusatives of limitation and of material respectively. However, I
maintain that both are better understood as arguments of their
respective verbs. It is the verb’s semantics that accounts for the
idiosyncrasy of the noun phrases it selects. When this is recognized the
pool of ‘adverbial’ or ‘modifier’ noun phrases is much reduced.
Consider the example from Joshua 7:12, the so-called accusative of
ל ְפ ֵנ ֣י ֹא ְיביהם ֥?י ה ֖יfi ל ֵח֑ ֶרם (6) ֗
ֹע ֶר ף י ְפנfi֙
‘They will turn (the) nape towards their enemies because they have
ḥērem.’ (Josh 7:12) (Waltke-O’Connor 1990, §10.2.2e)
In the paper, I argue that the Qal of פנה, though
most often monovalent, meaning ‘X turned,’ also has a bivalent pattern
selecting a Theme and yielding ‘X turned Y.’ I compare the root
פנה with other roots in the same semantic range,
such as הפך,
סבב, and וב?. My
conclusion is that the noun ֶרף
and not a modifier.
in Josh 7:12 is an argument
A similar argument is made for the noun זהב
occuring in Exodus 25:18.
ַנ֥ ִים ? ֻר ִ ֖בים ָז ָה֑ב (7) וע ִ
‘And you shall make two cherubim (out of) gold.’ (Exod 25:18)
(Arnold-Choi 2003, §2.3.2f)
If we accept that the Qal of
ה?ע has a trivalent
pattern meaning ‘X made Y out of
Z,’ then the material זהב is an argument selected
by the verb’s semantics. Therefore
we have no need for subcategories like ‘limitation,’ ‘material,’ or
‘instrument’ and category two becomes simpler and more coherent.
Meanwhile the wide range of noun phrases falling under category one are
explained by the equally wide range of verbal semantics which any
language must have.
## NP Secondary Predicates
Whereas both arguments and adverbials interact with the verb to describe
exactly what sort of event we are dealing with, secondary predicates
have a relationship to an argument and not to the verb. I adopt
Rothstein’s (2011, 1442) definition of secondary predicate with some
(8) A secondary predicate is a one place, non-finite predicate
expression which occurs under the scope of a main verb. Crucially, the
secondary predicate shares an argument with the main verb such that
the subject of the secondary predicate is either the subject or the
direct object of the main verb.
The most common types of secondary predicate mentioned in the linguistic
lit- erature are depictives and resultatives, which correspond roughly
to accusatives of state and product in the standard grammars. Depictives
and resultatives are similar in that they specify the state of one of
the arguments. They differ regarding when the state holds. Depictives
specify a state that holds while the event of the verb is ongoing, while
resultatives specify a state resulting from the event. Secondary
predicates may be subject- or object-oriented. For example, the English
sentence Jonathan*i* left the party angry*i*
contains the subject-oriented depictive angry. The sentence Moses
painted the lintel*i* red*i* features the
object-oriented resultative red.
Consider 2 Kings 5:2:
‘And Aram went out (as) raiders.’ (2 Kgs 5:2)
֣אfi ְגדfi ִדים י ְצ
The noun ִדים fiְגד ,
labelled an ‘accusative of state’ by Joüon-Muraoka, is a subject-
oriented depictive. Its function in the clause is to predicate of the
subject ָרם֙ א that
– at the same time that they went out – they were raiders. That is to
say, the Arameans went out and they were
a raiding party.
Similarly, consider 1 Kings 18:32:
את־האב ִנ֛ים ִמ ְז ֵ ֖fi ַח fi ֵ?֣ם יהו
֑ה (10) ו ִfiב ֶנ ֧ה
‘And he built the stones (into) an altar in the name of the lord.’ (1
The noun ַח fi֖
מ predicates of the object
ִנ֛ים את־האב that
when the building action
was complete, the stones were now an altar. That is, Elijah worked on
the stones and the stones became an altar.
Secondary predicates have their own distinct syntactic structure, which
I discuss in the paper. It is important that each category be defined in
clear syntactic and semantic terms to show why they are distinct.
4. # Three Types of Locative ‘Adverbial’ NP
6. ## An Event Semantics Approach to Modifiers
In this final portion of the paper I assume a Neo-Davidsonian event
semantics, which means simply that I assume that propositions can be
treated as introducing an ‘event’ or ‘eventuality’ into the discourse
that can be referred to and modified.
## The Syntax and Semantics of Locative Modifiers
I also follow Claudia Maienborn in distinguishing between modifiers that
apply to the whole event, to some part of the event, and to the entire
proposition. Each of these
three types has its own site of adjunction corresponding to its semantic
## Locative NP Modifiers in BH
What was traditionally called ‘adverbial’ should instead be referred to
by the more general term ‘modifier.’ The reason for this is that some
modifiers can be shown to modify the event as a whole – this is the
Event-External Modifier – while others mod- ify only a part of the event
by way of modifying the verb – this is the Event-Internal Modifier. The
former could be informally termed ad-eventive and the latter ad-verbal.
From now on, therefore, I restrict the term ‘adverbial’ to the latter
subcategory, the Event-Internal Modifier.
If we consider 1 Kings 8:32 again, we may note how
מים?ה specifies a
location for an event of hearing. There is nothing about the location
that affects what sort of event it is. It is an event of hearing whether
it takes place in heaven, in the temple, or in a closet. The phrase
מים?ה is therefore an
By contrast, there is the case of Genesis 3:15:
‘He will wound you (on the) head and you will wound him (on the)
heel.’ (Gen 3:15) (Waltke-O’Connor 1990, §10.2.2d)
This is admittedly a challenging case, and one might want to argue that
?ֹא ר and
ע are arguments in a trivalent pattern of the
verb. However, these nouns can be
understood as locative modifiers in the Event-Internal sense. The head
and heel are the surfaces of contact for the action of striking or
wounding. That is, the locations are part of the descriptions of these
two distinct events. The first is an event of wounding on the head,
which is not at all the same as the second, which is an event of
wounding on the heel.
Finally, consider אB֖
in Genesis 41:40:
ַ ֥fiל א ְג
‘Only (with respect to) the throne will I be greater than you.’ (Gen
Arnold & Choi label this noun phrase an ‘adverbial accusative of
specification.’ How- ever, it fits well with Maienborn’s third type of
locative modifier, the Frame-Setting Modifier. A frame-setting modifer
restricts the domain of a proposition to a specific location.
Maienborn’s test for frame-setting modifiers is to check whether
omitting the locative changes the truth value of the proposition. And
indeed, if we reduce the clause to flממ
אגדל ‘I am or will be greater than you,’ we see
that it is no longer true. In the context Pharaoh is asserting that
Joseph has the full authority of Pharaoh – that is, they are equals – in
every respect except for the throne. Only
Pharaoh has right to the throne and at the throne Pharaoh’s word
supersedes that of Joseph.
# 6 Conclusion
In summary, then, the heading ‘Category 2’ in Table 1 can be much
simplified. Instead of proliferating many idiosyncratic terms on the
basis of a noun’s semantics, it is more reasonable to restrict the
subcategories to ones that can be defined according to syntactic
structure and according to whether they modify the whole event, only
part of the event, or the entire proposition. The three subcategories
proposed here are Event-External Modifiers, Event-Internal Modifiers,
and Frame-Setting Modifiers. As it stands, Biblical Hebrew modifiers
seem to belong to three basic types – time, manner, and place – and
these correspond well with Larson’s findings for English.
The noun phrases considered here that are not modifiers have been
classified either as arguments or as secondary predicates. The wide
variation in semantics at- tested for ‘accusative’ noun phrases is
largely a result of verbal selection. Arguments are those phrases
required by the semantics of a verbal valency pattern.
Accounting for secondary predicates demands a third distinct category,
matched by distinct syntax and semantic representation. Secondary
predicates are, as the name suggests, a second predicate in a clause
which predicates something about one of the arguments of the verb,
either a subject or object. Noun phrases previ- ously labelled adverbial
accusatives of state and of product are in fact depictive and
resultative secondary predicates.
Arnold, Bill T., and John H. Choi. 2003. A Guide to Biblical Hebrew
Syntax. Cam- bridge: Cambridge University Press.
Bowers, John. 1993. “The Syntax of Predication.” Linguistic Inquiry 24
. 2001. “Predication.” In The Handbook of Contemporary Syntactic
edited by Mark Baltin and Chris Collins, 299–333. Malden: Blackwell.
Chomsky, Noam. 1993. “A Minimalist Program for Linguistic Theory.” In
The View from Building 20: Essays in Linguistics in Honor of Sylvain
Bromberger, edited by Kenneth Hale and Samuel Jay Keyser, 1–52.
Cambridge: MIT Press.
. 1995. The Minimalist Program. Cambridge: MIT Press.
. 2000. “Minimalist Inquiries: The Framework.” In Step by Step: Essays
on Minimalist Syntax in Honor of Howard Lasnik, edited by D. Michaels
and J. Uriagereka, 89–155. Cambridge: MIT Press.
Cook, John A. 2012a. “Valency: The Intersection of Syntax and
Semantics.” Paper Presented at the Annual Meeting of the Society of
Biblical Literature. Chicago, IL, November 17.
. 2012b. “Valency: The Intersection of Syntax and Semantics.” https :
/ / ancienthebrewgrammar . wordpress . com / 2012 / 12 / 05 / biblical
- hebrew - valency/.
. 2014. “Verbal Valency in Biblical Hebrew and the Case of
מלא.” Paper Presented at the Annual Meeting of
the Society of Biblical Literature. San Diego, CA, November 22.
Cuervo, María Cristina. 2003. “Datives at Large.” PhD diss., MIT.
. 2015. “Causation without a CAUSE.” Syntax 18 (4): 388–424.
Gesenius, Wilhelm, E. Kautzsch, and A. E. Cowley. 1910. Gesenius’
Hebrew Gram- mar. 2nd English ed. Oxford: The Clarendon Press.
Halle, Morris, and Alec Marantz. 1993. “Distributed Morphology and the
Pieces of Inflection.” In The View from Building 20: Essays in
Linguistics in Honor of Sylvian Bromberger, edited by Hale Kenneth and
Samuel Jay Keyser, 111–176. Cambridge: MIT Press.
. 1994. “Some Key Features of Distributed Morphology.” MIT Working
Papers in Linguistics 21:275–288.
Harley, Heidi. 2012. “Semantics in Distributed Morphology.” In
Semantics: An In- ternational Handbook of Natural Language Meaning,
edited by Claudia Maien- born, Klaus von Heusinger, and Paul Portner,
3:2151–2172. Berlin: Mouton de Gruyter.
Joüon, P. Paul, and T. Muraoka. 2006. A Grammar of Biblical Hebrew.
Roma: Pon- tificio istituto biblico.
Larson, Richard K. 1985. “Bare-NP Adverbs.” Linguistic Inquiry 16 (4):
595–621. Malessa, Michael. 2006. Untersuchungen zur verbalen Valenz im
Assen: Van Gorcum.
Rothstein, Susan. 2011. “Secondary Predicates.” In Semantics: An
International Handbook of Natural Language Meaning, edited by Klaus
von Heusinger, Claudia Maienborn, and Paul Portner, 2:1442–1262.
Berlin: De Gruyter Mouton.
Van der Merwe, Christo H. J., Jackie Naudé, and Jan H. Kroeze. 1999. A
Biblical Hebrew Reference Grammar. Sheffield: Sheffield Academic
Waltke, Bruce K., and M. O’Connor. 1990. An Introduction to Biblical
Winona Lake: Eisenbrauns.
Wilson, James Douglas. 2014. “Verbal Valency in Biblical Hebrew: An
Analysis of the Valency of עבר.” Master’s
thesis, Asbury Theological Seminary.