The volume is a collection of 13 articles selected from the first North American Conference in Iranian Linguistics (NACIL1). It presents articles on various components of the grammar of Iranian languages within different theoretical frameworks. Though the major focus of the articles is Modern Persian, historical (diachronic/ typological) aspects of Iranian languages are also included. I will group the articles sharing similar data or approaches together for review.
Iranian languages and the contents of the book are introduced by Moradi in chapter 1, “Advances in Iranian linguistics”. This is followed by Abodollahnejad and Storoshenko's analysis of xod-eš and un in chapter 2, “Syntactic and semantic constraints on pronoun and anaphor resolution in Persian”, which supports Kaiser et al.'s (Reference Kaiser, Runner, Sussman and Tanenhaus2009) multiple constraints framework, arguing that both syntactic structure and semantic information are considered to play roles in reference resolution.
Anonby, Hayes and Oikle's chapter 3 is a historical/typological analysis entitled “A multi-dimensional approach to classification of Iran's languages”, which highlights the shortcomings of two-dimensional models (the tree model and the wave model) of classification, and presents a multi-dimensional approach to the classification of Iranian languages in the form of a forced-directed graph. Their analysis is based on genealogical inheritance, structural similarities and association through ethnic identification.
In chapter 5, “The pronoun-to-agreement cycle in Iranian: Subjects do, objects don't”, Haig's diachronic study shows that subject pronouns, but not object pronouns, went through successive grammatical changes that resulted in their functioning as subject agreement morphology in the Middle West Iranian languages. He suggests that this distinction between subject and object is related to the informativity of the subject as opposed to the object, in that the former indexes person, whereas the latter signals gender and number.
A similar treatment of subject agreement morphology is also found in chapter 8, “Topic agreement, experiencer constructions, and the weight of clitics” by Jügel and Samvelian. Subjects are said to have appeared as hanging topics that were resumed by enclitic pronouns, before the latter were reanalyzed as verbal agreement markers, replacing the object agreement markers on the verb.
Chapter 4, “The additive particle in Persian: A case of morphological homophony between syntax and pragmatics” pits forward an insightful discussion by Ghomeshi on the additive -æm (hæm) as a pragmatic or post syntactic particle. Using the Callfriend Farsi corpus, she presents the forms, the semantics, and the environments where additive -æm appears. According to Ghomeshi (p. 70, 73), -æm may not appear in clause-final position nor on inflected verbs cross-linguistically. I would like to note, however, as shown in (1) below, that the additive may appear on an inflected verb, indicating her corpus was not large enough to include all syntactic categories that may host the additive.
(1) mi-xor-æm-æm qor mi-zæn-e
asp-eat-1sg-sg nag sg-hit-3sg
‘Even when I eat, he nags.’
Also, distributionally, the additive may appear on what seems to be a semantically vacuous third person singular pronominal enclitic not found in Ghomeshi's corpus.
(2) čamski-š-æm ne-mi-tun-e e'teraz kon-e
Chomsky-3sg-add neg-asp-can-3sg objection do-3sg
‘Even Chomsky cannot object.’
In chapter 6, “The suffix that makes Persian nouns unique”, Jasbi presents an in-depth analysis of the semantics of the suffix -e in colloquial Tehrani Farsi. He shows how the interpretation of nominals changes in different contexts when they are suffixed with -e. He presents a formal account in which this suffix contributes to a unique definite interpretation of bare nominals and restricts the domain of quantification to a singleton indefinite nominal that it attaches to.
In his second article, chapter 7, “The meaning of the Persian object marker rā:
What it is not, and what it (probably) is”, Jasbi analyzes ra as conveying old or presuppositional information, a common property found in previous treatments of this morpheme. He proposes a formal compositional account of objects marked with ra in Persian in which it implies the existence of the noun it modifies as part of the common ground in the conversation. He accounts for lack of ra on proper nouns as in (3) (p. 131: 24a) in the relevant context by saying that the existence of the entity Ali Saburi is not presupposed in that context.
(3) æli (e) Sæburi mi-šnas-i ?
Ali (ez) Saburi mi-know-2sg
‘Do you know anyone named Ali Saburi?’
Of interest, however, is that despite all the merits of Jasbi's account, it seems to fall short of explaining the lack of this particle on Hafez or Qur'an in (4), despite the fact that both may be assumed to be presupposed.
(4) Hæsæn qur'an/hafez mi-xun-e.
Hasan Quran/Hafez asp-read-3sg
‘Hasan reads the Quran/ Hafez.’
There is also a formal treatment of ra by Karimi and Smith in chapter 9, “Another look at Persian rā: A single formal analysis of a multi-functional morpheme”, analyzing it as a dependent Case marker which is assigned post-syntactically, a proposal Darzi (Reference Darzi2009) employed to resolve problems of Case and agreement in what he claims to be Subj-Subj-Raising constructions out of finite clauses. It marks Case on specific objects, on subjects raised to certain superordinate clauses, on topicalized DPs coindexed with a clitic inside an object, and on nominal adverbs in the domain where they are c-commanded by another DP, (e.g., the subject in the Spec of VoiceP above vP). They also extend their analysis to data from Modern Classical Persian.
A couple of points are in order here. First, in Karimi and Smith's analysis, a specific direct object may be assigned dependent case in spec,vP, to which higher adverbs may adjoin. Given this, the sentence in (5), in which a nominal adverbial not marked with ra is sandwiched between the direct object marked with ra and a higher adverbial, seems to pose a problem for their analysis. Under their analysis, the nominal adverbial is the higher nominal constituent in the domain of the subject and should, hence, be marked with ‘ra’. However, it is the lower direct object that is so marked.
(5) mæn xošbæxtane færda æli-ro mi-bin-æm
I fortunately tomorrow Ali-ra asp-meet-1sg
‘Fortunately, I will meet Ali tomorrow.’
Second, data not discussed in their analysis suggest that in the presence of a DP marked with ra in an embedded clause, an understood raised subject may not be marked with ra. If the embedded subject is merged in Spec TopP, which is an option in their analysis, it is unclear why it should be sensitive to the presence of a DP+ra in the embedded clause.
(6) æli(*-ro) fekr mi-kon-æm færda hæsæn-o be- šnas-e
Ali(*-ra) thought asp-do-1sg tomorrow Hæsæn-ra sbjv-recognize-3sg
‘I think Ali will recognize Hasan tomorrow.’
Note that Karimi and Smith analyse the DP Ali in (6) without ra/ro as a base generated topic that does not originate in the embedded clause. Now, given the clausemate requirement between a negative concord item (hičkæs ‘nobody’ in (7)) and a negative marker in Persian, the grammaticality of the sentence in (7), in which the negative polarity item is in the main clause while the embedded verb is negated remains a mystery, in their analysis. This sentence shows that in the grammatical version of (6), Ali may not be treated as a base generated topic, but rather as having moved there.
(7) hičkæs fekr mi-kon-æm færda hæsæn-o næ- šnas-e
nobody thought asp-do-1sg tomorrow Hæsæn-ra neg-recognize-3sg
‘I think nobody will recognize Hasan tomorrow.’
Third, it is unclear, under their analysis, why the specific direct object hafez, both in (4) above, and in (8) below where it has raised into the higher clause, may appear without ra.
(8) mæn hafiz dust dar-æm [CP PRO be-xun-æm]
I Hafez like have-1sg sbjv-read-1sg
‘I like to read Hafez.’
The last contribution on ra belongs to Suleymanov; “Oblique marking and adpositional constructions in Tat: A mosaic of dialectal convergence and divergence” (chapter 13), which discusses the function and distribution of r(A) in Tat in different varieties and regions. He notes that other markings on objects, whether they are inherited or innovated, are due to language contact that resulted in the extension of the function of this morpheme. Different varieties of Tat are shown to behave differently with respect to marking both direct and indirect objects with r(A) in ditransitive constructions. The study presents similarities and differences among different varieties of Tat in two different possessive constructions as well.
In chapter 10, “The Ezafe construction revisited” Larson and Samiian, building on Samiian’'s (1994) analysis of ezafe as a Case assigner, account for the presence of ezafe before reduced relative clauses (data not accounted for under previous analyses). They propose that ezafe intervenes between two adjacent [+N] categories to assign Case to the following nominal element (p. 199). This also applies to non-finite relative clauses and infinitival complement clauses in Persian, as opposed to Sorani and Kurmanji. They show that (i) ezafe is required before propositional XPs bearing a nominal feature, and (ii) ezafe attaches to propositional XPs if their final element is nominal. They posit a pP above some Persian PPs, and a nominalizing morpheme 
$\surd n$ in different positions in Class 2 prepositions that converts the categorical status of the PPs into NPs. Ezafe would then be responsible for providing Case for the NPs/DPs in Class 1 and Class 2 Ps, while the p is responsible for providing Case for the DP complement of P in Class 1 Ps.
Larson and Samiian seem to underestimate cases where ezafe appears on a noun taking a Class 1 P by saying that in the 126 cases of Class 1 P in the corpus they used, only one instance of such a string was found (p. 220, fn. 27). However, one can think of many strings like the one in (9), where a Class 1 P is preceded by ezafe. This, together with the data I introduced in my critical review of other articles in this volume, makes it clear that a corpus, large as it may be, may not always give a full picture of the facts.
(9) Jiq *(-e) ba seda-ye bolænd-e u
scream *(ez) with voice-ez loud-ez her
‘her loud scream’
In chapter 11, “Quantitative meter in Persian folk songs and pop lyrics”, Mahdavi Mazdeh shows the empirical and theoretical problems with previous analyses of the metrical system of folk songs and pop lyrics in Persian. He argues that this system is quantitative and follows the same principles found in Classical Persian metrics, with the minor differences between the two mainly having to do with optional vowel shortening in colloquial Persian.
Chapter 12 is devoted to Rasekhi’'s analysis of “Stripping Structures with Negation in Persian”. She presents a tripartite classification of stripping constructions as Polarity Stripping (PolS) and Negative Stripping (NegS) which involve clausal coordination and TP ellipsis, and Pseudo-stripping (PseS) which is mono clausal and involves movement.
In sum, the volume offers novel approaches to well-known topics such as ra and ezafe, as well as discussing some novel and less well-studied topics within current theoretical frameworks. It contains valuable articles on different topics that shed light on aspects of Iranian languages. In my view, the articles in this volume provide good materials for graduate-level seminar courses on Iranian languages and will definitely be appreciated by scholars of Iranian linguistics.
