Filipino Children’s Acquisition of Nominal and Verbal Markers in L1 and L2 Tagalog

Round 1

Reviewer 1 Report

Dear Author(s) of the manuscript "Filipino children’s acquisition of nominal and verbal markers in 2 L1 and L2 Tagalog"!   I read carefully the manuscript and, unfortunately, found it absolutely irrelevant from the perspective of formal requirements stated in the MDPI Instructions for Authors. E.g. the structure of the manuscript is not relevant neither to the structure of a research paper or review paper; citations are not precise (some of the citations do not correspond to the original texts; some of the cited papers are not possible to find in any database); statistical analysis is missing. Although the topic and issues discussed in the manuscript are interesting and need scientific research, the manuscript, at this stage, can not be published. Please, read carefully the MDPI Instructions for Authors and rewrite your manuscript accordingly.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This is a review paper that sums up the work on the development of noun and verb inflection in Tagalog to date. It is an extension of Gonzalez’s (1986) review, but with the goal of presenting a fuller picture of the commission errors produced by Tagalog-learning children, both L1 and L2 learners.

The paper is by its nature descriptive, rather than addressing key theoretical questions, but I think it is interesting, nonetheless. That said, I do think the authors could speculate more about the theoretical implications of this work. My main points are below:

 Major points:

 1.      First section of the introduction: This is a nice summary of some of the literature on errors but it is very narrow; it focusses on questions and the past tense errors, rather than summarising (succinctly of course) the much wider cross-linguistic literature on the acquisition of verb (tense, person, number) and noun (case) inflection. I would also like to see more discussion of the theoretical implications of these debates. A good place to start to find this literature would be the chapter on inflection by Ambridge & Lieven’s (2011) book Child Language Acquisition. There have also been some very nice cross-linguistic papers coming out of those authors’ labs, which are a good reference for more up to date work.   This doesn’t need to make the introduction substantially longer.  I would actually remove the paragraph on question errors since those aren’t the topic of the paper and focus more on the literature on verb and noun morphology, and summarise that literature more succinctly.

 

2.      Method: I would like to see more details of the screening process and the criteria that the authors used to include/exclude papers. This is important so that the readers can be confident that no papers were missed.   I realise that the authors did not use a full ‘systematic review’ process but it still would be useful to look at one of those processes, and to identify in more detail which steps they went through (see e.g. https://lib.guides.umd.edu/SR/steps). For example, what databases did they use for their search - both for published and accessible non-published work, what keywords did they enter, what were exclusion and inclusion criteria, did they first do a title-based exclusion process, then an abstract based process etc...

3.      Method: Similarly, it would be good to provide a flow diagram detailing the steps, and how many articles were included at each step.

4.      Method: the same comment as 2 above goes for the dataset search. How did they identify the relevant datasets, how sure are they that they didn’t miss any?

5.      Results: The authors finish the first two subsection on L1 acquisition, and the second on L2 with a short summary of the findings. It would be useful to do this with the L2 findings about verb nominal markings too.

6.      Results: Page 8: paragraph starting “Most of the studies reported...”.  This paragraph discusses the fact that most errors occur in structured tasks, and suggests that this might be because such tasks are more complex. However, an equally likely explanation is that error rates tend to be higher in less frequently produced structures, which are (by definition) less likely to occur in naturalistic speech (see Aguado-Orea & Pine, 2015, PLoS One). Thus, we only discover these errors when we use structured tasks.

7.      Discussion: Above, I suggested that the authors spend more time in the introduction discussing the theoretical implications of the debates in the literature on errors in the development of noun and verb inflection.  I would like to see this mirrored in the discussion; with a short section on what the authors think the pattern of errors in Tagalog might mean for these debates, and how future work might test the predictions of the different theories using Tagalog.

 Minor points:

8.      I wasn’t sure why the authors used the term ‘non-realis’ instead of ‘irrealis’. Is the meaning different?

9.      Page 4: There is a short discussion of potentive and stative verbs. It wasn’t clear to me what markers are used for these? Is it maka- (agent voice) and ma- (patient voice) for potentive, and naka (agent ovice) and n- (patient voice) for stative? Can this be made clearer?

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Thank you for the opportunity to review this manuscript. This article provides a review of the production of commission errors among L1 and L2 learners of Tagalog, a Western Austronesian language with complex voice systems, based on experimental and observational studies. The results of their review showed that Tagalog L1/L2 learners make errors of commission on verbal morphology and nominal marking, with an observable decrease in the frequency of these errors over time. The review is a valuable contribution in the literature on Tagalog language acquisition. However, I would like the authors to address the following major and minor issues.

Major comments

1.  On cross-linguistic generalizability and theoretical implications: I do understand the paper was intended to do a descriptive review on Tagalog, but I would like to encourage the authors to expand on their discussion and discuss whether the findings from previous Tagalog studies are consistent with findings on other languages (i.e., errors on verbal morphology and case marking, which are linked to problems with syntactic functions), and how the data, in general, inform us of our understanding of commission errors in language acquisition. I don’t think the authors need to adhere to a specific theory or hypothesis, but a discussion on the potential implications of these findings would be relevant in (re)thinking about errors of commission in language acquisition. I think incorporating this will benefit the article in reaching not just Tagalog or Philippine language researchers but a wider research audience. Some of the comments/questions below also focus on situating the review in the larger research context on language acquisition.

Further, I am interested if the authors could also include in their review if there are observable patterns reported from the studies that could further shed light on the acquisition of this voice system:

a.  Are there more frequent error rates involved with the subject marker ‘ang’ (i.e., production of the double-ang marker, as the authors call it) or the nonsubject marker ‘ng’ (i.e., double-ng)?

b.  Were there more dominant patterns of commission errors in one voice pattern (patient voice) versus another (agent voice)? While there are methodological differences across studies, there seems to be patterns observable across studies, as mentioned in Garcia et al. (2018), Garcia (2021), and Galang (1982).

Can the authors also comment in the Discussion section on what the possible source(s) of this difficulty is with the voice system, that leads to the production of commission errors?

2.  Description of the Tagalog voice system. On page 3, line 123 (and beyond, as applicable), the authors use the term “patient voice” label for all non-agent voice patterns. Could the authors explain this choice of terminology for all non-agent voice patterns? I understand that oversimplifying the voice system could benefit readers who are not familiar with the language, but it might compromise a more accurate description of the complex grammatical system of the language.

a.  While many of these adjunct-like thematic roles could be subsumed under a macro proto-patient category, there are designated affixes for the other kinds of thematic roles in Philippine-type languages, such as location, instrument, or benefactive, and they are often described in the Western Austronesian voice literature with the more specific labels such as locative voice, benefactive voice, instrument voice, conveyance/circumstantial voice, etc. There is extensive work in formal and descriptive syntax that do not lump the non-agent voice patterns under a single category “patient voice,” as there are designated affixes in Tagalog (as well as in other Philippine-type languages) for patterns in which the “adjunct-like” thematic roles (e.g., locative, instrument) are marked with ‘ang’ (e.g., Riesberg, 2014).  While work in language acquisition/processing remains scarce with these voice patterns, there is no principled reason to place them under a single category, as these non-agent voice patterns might not be acquired/processed in the same way.

b.  There is always no one-to-one mapping between the so-called “voice” and the “thematic role” chosen as the subject, so to make a distinction between agent voice versus non-agent voice (and using the label “patient voice”) could be problematic. For example, some “agent voice” affixes do not always map to an agent – some verbs marked with the “agent voice” could take experiencer subjects, or theme subjects (in the case of unaccusatives; see Chen and McDonnell’s (2019) review).

I would recommend the authors (i) revise the labels to appropriately identify the affixes based on how they are more conventionally used, i.e., by the thematic roles these affixes are more often associated with (e.g., -an for the locative voice, i-/ipag- for benefactive voice) or (ii) restrict the discussion to the conventional description of agent and patient voice, since the majority of the examples and studies do focus on these two common voice patterns.

 

Line-by-line comments

p. 2, lines 77-85: Briefly elaborate the significance of Tagalog in highlighting our understanding on commission errors.

p. 3, line 99: Clarify: Humihila NG babae, not ANG babae?

p. 3, line 131: Clarify: in the agent voice, isn’t the -um- present in the non-realis perfective mood?

p.5, line 182-184: Also mention the possible advantage of this review to other Philippine (or Philippine-type) languages with the same voice system?

p. 5, line 191-192: I would be interested to know the raw frequencies of the studies found in the search.

p. 5, line 213: Edit: 2;1, not 2.1

p. 6, line 236: gloss for ‘ako’ should be ‘1SG.SBJ’

p. 6, line 244: Related to the comment above on oversimplification of labels: Consider having two separate glosses for the nominal markers ‘ng’ and ‘sa’? While both are non-subjects, ‘ng’ could either be NSUBJ or GEN (genitive), while ‘sa’ tends to be used for oblique goals or locatives.

p. 7, line 296: Edit: shouldn’t the marker for ‘babae’ be ‘nung (and not ‘yung), for the NSBJ?

p. 7, line 306-317: Based on the Garcia et al. (2018) study, I wonder if the greater number of errors between the agent vs patient voice could be accounted to frequency – because patient voice tends to be more frequent in the language, hence commission errors are less frequent in the PV than in the AV?

p. 9, line 418-419: Clarify: Aren’t these three instances considered correct, for the subject to be marked with ‘ang’ and for the non-subject to be marked with ‘ng’?

p. 10, line 462-463: For the ease of comprehensibility for the readers who are less familiar with the voice system, could the authors also briefly expound on why there is a negative transfer from L1 Chabacano to L2 Tagalog?

p. 14, line 649: Edit: ‘marker’, not ‘maker’

References

Chen, V., & McDonnell, B. (2019). Western Austronesian voice. Annual Review of Linguistics, 5, 173-195.

Riesberg, S. (2014). Symmetrical voice and linking in Western Austronesian languages. In Symmetrical Voice and Linking in Western Austronesian Languages. De Gruyter Mouton.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Please see attached PDF file

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear Authors of the manuscript! I accept all improvements and additional information and I would accept the manuscript for publication. A few minor changes still could be recommended:

1) The aim of the study could be stated more explicitly in the body of the text.

2) The text in Fig. 1 could be written using a larger font size.

Sincerely,

 

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have addressed all my comments. I congratulate them on a very nice paper. I also very much like the addition of table 2, which makes the findings very clear. I have only a few stylistic comments remaining:

Page 1, line 21: replace 'chlidren's life' with simply 'life'.

Page 2, line 68, replace 'less stem changes' with 'fewer stem changes'

Page 3 footnote - the link to this comes in the middle of example 1 and I spent quite a lot of time looking for this. perhaps it could come at the end of the sentence 'Pronouns also have different ang, ng, and sa forms' instead

Page 6 line 244: replace 'wrote' with 'writing'

Page 9 line 366: replace 'was omitting one' with 'concerned omitting one'

Page 10 line 417: replace 'on Table 2' with 'in Table 2'

Page 10 line 429: replace 'older age range as Gonzalez' with 'older age range to that of Gonzalez'

Page 10 line 436-7: I found it hard to parse the meaning of the phrase 'as much as this is only expected in root words....' Can this be rephrased?

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 3 Report

 

I was a previous reviewer of the manuscript. I am pleased with the revisions and the responses made by the authors with the manuscript, and I recommend acceptance. I have additional comments below, which I would suggest the authors to incorporate in the final form of their manuscript.

1.       I recommend that the authors add their response to question 1a (in the response to reviewers) as a footnote:

“Considering the data from both experiments in Garcia & Kidd (2020), it does not seem to be the case that there were more double-ang than double-ng errors in general. However, it seems that there were slightly more ang-errors in the patient voice than in the agent voice (Exp. 1: 5% of total productions in AV were double-ang, while it was 10% for PV; Exp. 2: 4% double-ang in AV, 6% in PV). Then again, the bulk of the errors in the agent voice were reversal errors, i.e., use of ang instead of ng and vice versa; and overall, there was also higher accuracy in noun-marking in the patient voice than in the agent voice. In L2, there were considerably more double-ang errors by the Chabacano-speaking children than their Cebuano comparison group.”

2.       I think that there are sections where the authors could strengthen the connection between the theoretical accounts mentioned in the Introduction and in the Discussion. I suggest that authors add some brief statements in the following lines/parts of the manuscript that mention how looking into Tagalog could also inform our understanding on the acquisition of inflectional morphology, or on the nature of commission errors.

p. 2, line 96

p. 5, line 197

3.       Minor edits

-          Line 23: should be 1), not a) for omission errors

-          Page 3, footnote 1: SBJ for subject, GEN for genitive, DAT for dative (not ‘to’)…

-          Line 144: Delete ‘and’ and ‘aspect’ in ‘non-realis and imperfective aspect;’ I think it should be 'non-realis imperfective' to refer to a single category in the 2x2 cross between mood and aspect

-          Please double check the consistency of the abbreviations for the glosses, e.g., SBJ, not SUBJ. Multiple inconsistencies are observed throughout the manuscript.

-          Gloss in line 277, sentence (7a): gloss for ‘sa’ should be DAT, not NSBJ

-          Gloss in line 645, sentence (27): gloss should be POTPV

I like the addition of Table 2. Some suggested edits:

-          Clarify the scope of the work for the caption in Table 2. It should read as: 'Summary of L1 and L2 errors in Tagalog reported in previous work'

-          For Table 2, I suggest the authors also organize (by labeling the rows and repositioning contents) by categorizing the error types in terms of voice morphology/nominal marking. Also, shouldn’t Table 2 be positioned after Section 3.2.2, after discussion of L2?

 

 

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 4 Report

The manuscript is improved from the previous version. I have two minor comments that I think would be good to address, but I leave it up to the editors/authors to make a decision.

First, the authors have added a footnote (fn 4) in response to my comment about the finding regarding MLU. However, I think the original point still stands. Why is it noteworthy that 80% of verbs produced by children with an MLU in morphemes of under 2 were uninflected? If the MLU in morphemes of a child is under 2, a majority of the utterances made by that child should have only a single morpheme. It should be likely then that most of the verbs uttered are monomorphemic (i.e. uninflected). For example, if 80% of the utterances have only one morpheme, then it should not be surprising or interesting that around 80% of the verbs uttered are also only one morpheme.

Considering this, my question is the following: why is it noteworthy that 80% of the verbs for children with an MLU of below 2 are uninflected? Is the increased rate of inflection for children with MLU > 2 expected? Higher than expected? Lower than expected?

Second, the summary table is a good start for organizing the results discussed in a clear manner, but I think it can be made a bit more succinct or at least easier to read. In the current revision, the table seems like a set of lists forced to fit into a table. If a table turns out to not be the best way to present a summary of the results, then the authors should find a different way of doing it.

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf