LINGUIST List 23.4339

Wed Oct 17 2012

Review: Language Documentation; Phonology; Morphology: Van Otterloo (2011)

Editor for this issue: Monica Macaulay <>

Date: 17-Oct-2012
From: Michael Marlo <>
Subject: The Kifuliiru Language, Volume 1: Phonology, Tone, and Morphological Derivation
E-mail this message to a friend

Discuss this message

Announced at
AUTHOR: Karen Van OtterlooTITLE: The Kifuliiru Language, Volume 1SUBTITLE: Phonology, Tone, and Morphological DerivationPUBLISHER: SIL International PublicationsYEAR: 2011

Michael R. Marlo, Department of English, University of Missouri


At over 1000 pages and with over 1300 linguistic examples, the two-volume work,“The Kifuliiru Language”, by Karen and Roger Van Otterloo, is an impressivelylarge and thorough linguistic description, based on many years of work with theFuliiru-speaking community, including over 13 years in which the authors livedamong the Bafuliiru in the Democratic Republic of Congo (formerly Zaire). Thefirst volume (on which this review focuses), by Karen Van Otterloo, details thephonology and tonal system of Fuliiru and also contains detailed informationabout aspects of the morphology of the language. The second volume, by Roger VanOtterloo, deals with primarily non-phonological grammatical and discourse-levelphenomena. Each of the two volumes is accompanied by an excellent preface byDerek Nurse, and a moving foreword by representatives of the Bafuliiru community.

Following a brief initial chapter, “Conventions”, Ch. 2 “Phonology” describesthe non-tonal phonology of Fuliiru, which includes the phonemic inventory of thelanguage, information about the phonological and morphological distribution ofthese phonemes, and syllable structure. Phonological processes identified andexemplified include rules of flapping, post-nasal voicing, nasal placeassimilation, vowel hiatus resolution, vowel lengthening, vowel shortening,vowel harmony, glide epenthesis, and spirantization. Ch. 2 also includesextended discussion of the phrasal cliticization of monosyllabic words (a topicaddressed again in the appendix).

The Fuliiru tonal system is the focus of the 175-page Ch. 3. As in most otherBantu languages (see Downing 2011, Kisseberth & Odden 2003), tone is lexicallycontrastive in Fuliiru and also has important grammatical functions. VanOtterloo analyzes Fuliiru as having three underlying tonal values: /H/, /L/, and/Ø/ (“toneless”). Underlying Hs typically surface [H]; underlying Ls typicallysurface [L]. Underlyingly toneless tone-bearing units (TBUs) can becomeaffiliated with H in the course of tonal derivations, in which case they surface[H], but if no other tone links to the toneless TBU, it surfaces [L]. Thus, L isanalyzed as the ‘default tone’. An interesting feature of Fuliiru tonology isthat many morphemes have the opposite tonal value compared to Proto-Bantu, e.g.the verb root *-gùd- is /L/ in Proto-Bantu but /H/ -gúl- in Fuliiru, while the/H/ root *-tég- ‘set (trap)’ of Proto-Bantu is /L/ -tèk- in Fuliiru.

Verb roots come in three tonal types -- /H/, /L/, or /Ø/ -- and all verb formsare inflected with a grammatical tonal suffix (“melody”), which is one of themarkers of tense-aspect distinctions on the verb. Two other types of morphemescontribute H tones to verb forms: human singular object markers and thecausative and passive suffixes. Melodic tonal suffixes, which are selected as anexponent of tense-aspect-mood inflection, are classified into two types: (i)‘simple’ melodies consisting of a single /H/ or /L/ melodic tone, and (ii)‘complex’, two tone melodies /HL/, /LH/, /LL/, and /HH/. The underlying floatingtonal suffixes typically link to the leftmost free TBU in the verb stem. In thecomplex tonal melodies, the first tone of the melody links toward the left edgeof the verb stem, while the second tone of the melody links toward the rightedge of the verb stem.

Monomoraic and bimoraic noun stems are divided into two basic tonal types: H vs.L. Most (but not all) longer nouns have either the H or L patterns or one of twotwo-tone patterns -- LH or HL -- whose tones are analyzed as melodies that areassigned to the noun using the principles established for verbs.

Ch. 4, “Derivational processes,” provides a detailed investigation intocategory-changing morphology in Fuliiru, including examples of nouns derivedfrom verbs, adjectives, other nouns (by changing the noun class prefix), andfrom ideophones, along with a handful of examples of V+N and N+N noun compounds.We also find verbs derived from nouns, ideophones, and adjectives, as well asexamples of the many verbal auxiliaries of Fuliiru that are derived from otherverbs. Finally, examples are provided of adjectives and adverbs derived from verbs.

Ch. 5, “Verb stems,” focuses specifically on the morphology of the verb stem inFuliiru. The productive and non-productive verbal suffixes of the language areidentified and exemplified, and the various suffix combinations encountered inthe author’s corpus are summarized here.

The appendix, “Determining word boundaries and related orthography issues”,describes the phonological and morphological diagnostics that the author uses indetermining word boundaries in Fuliiru. These include phonological properties,such as the presence of surface long vowels, tonal contours, and the outcome ofvowel hiatus resolution. Grammatical features used to diagnose word boundariesinclude the presence and placement of specific morphemes like subject markers,the ‘prefinal’ suffix -ag-, and grammatical tonal suffixes.


A considerable amount of important linguistic territory is covered in this book,which is the first significant description of the Fuliiru language. Thedescription of morphology in Ch’s. 4 and 5 is especially rich, and is possibleonly after long, sustained work with a language. The section on the tonal systemis noteworthy for its methodical treatment of the grammatical tonal melodies,showing how the melodies combine with the different lexical tonal classes ofverbs and with the causative and passive suffixes. Interesting data and strongevidence are marshaled concerning the complex concept of ‘wordhood’ in theappendix. (In fact, in my view, much of the information in the appendix is soimportant that it should have been incorporated into the main text.) Thethorough description of grammaticalized auxiliary verbs is also impressive.

Nonetheless, a number of recurrent problems, particularly in the presentation ofthe material, prevent this work from reaching its full potential. Thedescription and analysis are difficult to follow and evaluate in many places,some key content is missing or underrepresented, and important literature onBantu languages has not been cited. In my criticisms below, I begin by focusingon problems with the presentation; I then discuss missing content, issues withthe analysis, and finally problems involving citations.

Presentation issues: One of the main problems of presentation throughout Vol. 1is that the reader is often not sufficiently prepared to deal with specificfacts or analytical details when they are discussed because s/he needs toalready know information that is not presented until later. In many cases, whenthe relevant issues are addressed later, the information from the earliersections is not appropriately integrated into the discussion, making the latersections challenging to comprehend as well.

A case in point is the whole of Ch. 4, which systemically describescategory-changing morphology, including nominalizations. While many of thepatterns discussed in this chapter will be familiar to Bantuists, in order tounderstand this material, the reader needs to know about the noun class systemof the language and the morphological structure of nouns and verbs -- topicswhich are not covered until Vol. 2. I found no cross-references within Ch. 4 toany of the other locations in the book where morphology is discussed: Ch. 2 ofVol. 2 for noun morphology; Ch’s. 3 and 5 of Vol. 1 and Ch. 9 of Vol. 2 for verbmorphology.

Problems with the flow of information also play out at a low level in thepresentation of data. There is a tendency for large chunks of data to be givenat once, followed by all of the description of those forms. This formatunnecessarily requires the reader to constantly flip back and forth across pagesto compare the data against their description. A better format would split updata displays into relatively small pieces and deal with those piecesone-by-one, integrating description, data, and analysis, and then providingextended discussion as necessary.

In Ch. 3, the tone rules are systematically separated out from the discussionand analysis of the tonal data and are found over 100 pages into the chapter. Inthe earlier sections of description and analysis, the reader has to flip farahead in the book to figure out what some new rule does, but the later sectionitself is difficult to evaluate since the data that putatively motivate the ruleare no longer at hand. A more successful strategy would be to describe the dataand develop the analysis in tandem, presenting the rules motivated by the dataunder discussion, and then later summarizing the analysis and rules. A similarproblem is found in the treatment of monosyllabic verb stems, which are omittedfrom initial sections where longer verb stems are treated and treated separatelyin a later section. In this format, the reader cannot evaluate the analysis ofany sections describing a specific tonal melody without data from one of the keyshapes of verb stems. The section on the monosyllabic verb stems is alsochallenging to comprehend because the comparable forms from longer stems withthe same melody are not repeated.

Ch. 2 could be improved if the presentation of phonological rules werereorganized more coherently around phenomena such as ‘NC effects’, ‘vowel hiatusresolution’, ‘imbrication’, ‘minimality effects’, etc., which are theoreticallyinteresting phenomena that are well known and widely discussed in the literatureon Bantu languages. In the current format, many rules are organized randomly andineffectively. For example, although all of the rules that are necessary tounderstand vowel hiatus resolution are found in § “Phonological rulesaffecting vowels”, two unrelated rules involving vowel devoicing and/u/-deletion after /m/ intervene between rules dealing with /V+V/ combinationsand two processes intimately related to hiatus resolution: compensatorylengthening and post-glide vowel lengthening. Later within this section, we findlengthy remarks on a vowel lengthening rule involving clitics, which has nothingto do with vowel hiatus resolution, but which precedes a vowel shortening rulethat interacts with vowel hiatus resolution processes.

A section on morphologically governed rules is organized around rule types, e.g.vowel harmony, epenthesis, avoidance of N+N sequences, etc. This organizationputs together rules from diverse and largely unrelated contexts that generallydo not interact with one another, and it separates related and interacting rulesfrom one another. One regrettable consequence of this organization is that arule that inserts [i] between the 1sg nasal subject prefix N- and followingnasals is found in the section on rules of ‘epenthesis’ instead of the sectionon the avoidance of N+N sequences.

Content gaps: The interactions between phonological rules are generally notdiscussed anywhere, which is a significant omission given the importance of ruleinteraction to the development of any phonological analysis. A case showing thisconcerns a rule, common in Bantu languages, that lengthens vowels before NCsequences. This rule applies even when nasal segments are deleted before othernasals, but this fact is not mentioned in the dedicated sections on pre-NClengthening or pre-nasal nasal deletion. In a later section on a (possiblydiachronic) rule of ‘nasal effacement’, it is noted that pre-NC lengtheningapplies even when the nasal has been deleted before a voiceless fricative, but asecond rule interaction in the same data set is not mentioned: spirantizationtriggers the subsequent application of nasal effacement.

In Ch. 3, a wide variety of stem sizes and shapes are presented for theinfinitival forms of verbs, and basic paradigms are given for the many tonalmelodies in the language. However, not all factors affecting tone aresufficiently described in subsequent sections, where we do not find exampleswith verb roots longer than CVC or CVVC or examples of verbs combined withobject markers in the HL, LH, or LL melodies. Longer verb stems are oftencrucial in seeing the essential properties of a tonal pattern, and objectmarkers often trigger tonal alternations, so such data need to be systematicallypresented as part of the basic documentation and description of the tone system.

The treatment of Fuliiru phrasal tonology is a major weakness of Ch. 3. Thediscussion of the tonal patterns of phrasal forms is limited to the brief §3.12,over 160 pages into Ch. 3, after virtually all of the analysis of the tonalsystem has been presented. As with long verb stems and object markers,phrase-medial forms are one of the key contexts that must be treatedsystematically in the study of a Bantu tonal system. The few phrasal forms thatare provided suggest that data in phrase-medial position are essential tounderstanding tone in Fuliiru. For example, in §3.12, we learn of phrase-finalvs. phrase-medial alternations such as:

á-ká-shúlíkàhe-fut-hit‘he will hit’


á-ká-shúlíká múzíbò‘he will hit Muzibo’


úkú-kùúndàinf-love‘to love’


úkú-kùùndá múzíbò‘to love Muzibo’.

These forms bear on one of the major themes of Ch. 3: the “extratonality” of thefinal syllable of words, which is listed as one of the “basics” of Fuliiru tone(p. 111) and is used to explain at least two key properties of verb tone: (i) Htone does not spread onto the final syllable of the verb and (ii) grammaticaltone suffixes (usually) do not target the final syllable of verb stems.Extratonality figures into the analysis of virtually all of the data in thefirst five sections of Ch. 3 and is discussed separately in its own section,§3.6. As Ch. 3 concludes, we discover from phrase-medial forms like those abovethat tones do in fact spread to the stem-final syllable in phrase-medialposition. Unfortunately, so little phrasal data is provided that it is difficultto assess the descriptions and proposed analyses or evaluate reasonablealternatives, such as the possibility that the phrase-medial forms reveal thebasic tonal pattern and the phrase-final forms undergo rules. Other keyquestions also cannot be answered by the given data and discussion, such aswhether the stem-final syllable can be targeted by the tone assignment ruleswhen the verb is in phrase-medial position. If the answer to that question is“yes”, then much of the mechanics of the analysis presented in Ch. 3 needs to bereconsidered, since the discussion leading up to this point has repeatedlyasserted that the final syllables of nouns and verbs are invisible to toneassignment.

Analytical issues: Some specific analytic choices made in Vol. 1 are not welljustified. For example, we learn in Ch. 2 that the vowels /u o/ behave as anatural class with respect to a few phonological rules in Fuliiru, including twoof the vowel hiatus-resolving rules. The author always refers to /u o/ as the“back” vowels. The low vowel /a/ is excluded from this characterization becauseit is considered to be a “central” vowel. The possibility that the phonologicalfeature [+round] is responsible for the patterning together of /u o/ to theexclusion of /a/ is never mentioned. This is relevant because in §,it is crucial to the analysis that /a/ be treated as a third category ofbackness distinct from /i e/ and from /u o/. The need for a three-wayphonological distinction in backness goes away if rounding is taken into account.

The analysis of phonetic mid tone given in Ch. 3 and discussed in some detail in§3.3 is questionable. The distribution of mid tone is restricted such that itoccurs only after H and always derives from intermediate σ ́σ̌ (H-LH). In fn. 30(p. 121), the author mentions the possibility of analyzing the mid tone as adownstepped H, which would be the result of the L delinking and acting as adownstep operator on the following H -- the standard analytical move (Bird 1966:135, Clements & Ford 1979: 206, Paster & Kim 2011). The author even points out afurther argument for treating the mid tone as a downstepped H: the mid tone isphonetically identical to the automatic lowering of H after L through downdrift.Inexplicably, the author rejects the analysis of the mid tone as a downsteppedH, claiming that “in Fuliiru the L tone (as well as the H tone) of atautosyllabic LH is always linked to the TBU on which it is realized” (p. 121).The theoretical assumptions concerning the relationship between phonologicalrepresentations and phonetic forms underlying this rejection are unclear; thereis no further discussion of how the phonological representation in which a LHsequence linked to a single syllable translates into the phonetics as a mid tone.

Citations: The Bantu linguistics literature is only superficially integratedinto Vol. 1. In light of the fact that phonology and tone are the two mainsections of Vol. 1, it is striking that the work of several prominent Bantuistphonologists of the autosegmental era is missing or severely underrepresented inthe references. Key omissions include Bickmore (2007), all works by Downing,most of the relevant publications of Kisseberth and Odden, Hyman (2001), andStevick (1969). While some appropriate references to other languages with threetonal classes of verbs are provided, §3.3 does not give much contextualizationof Fuliiru’s noteworthy status as a ‘reversive’ language.

It is claimed in the book’s advertising that “these volumes aim at a thoroughpresentation of the many interesting features found in a typical InterlacustrineBantu (J) language”. In light of this claim, the almost complete lack ofinteraction with the literature on Bantu languages from Zone J is even moresurprising. Nothing is said about several J Zone languages that are the subjectof rich descriptions and analyses (especially with respect to tone): Haya, Jita,Kerewe, Bukusu and other Luyia varieties, Nande, Nkore, etc., and there is nomention of Harjula’s (2004) grammar of Ha, which, like the current two-volumeset, is based primarily on data drawn from texts.

The Bantu linguistics literature is also poorly represented on the issue ofprosodic minimality effects. This is a fairly minor omission, but the apparentlack of consultation of this relatively vast literature (see e.g. Downing 2001,2006, Park 1997) leads to an inaccurate characterization of this literature (p.425): “Polymorphemic words [in Bantu languages] are the rule rather than theexception. This fact is widely recognized. What is not so often noted is thatmany, if not most Bantu languages seem to have some degree of phonologicalconstraint against independent monosyllabic words.” In fact, such constraintsare well known and well discussed.

Conclusion: Vol. 1 of “The Kifuliiru Language” is a rich resource on aninteresting language that fails to satisfy the high expectations demanded ofdescriptions and analyses of the phonology, morphology, and tone of a Bantulanguage. In view of other high-quality works on Bantu languages and the depthand breadth of the corpus that the author has developed over many years ofresearch on Fuliiru, a more rigorous treatment of the facts of Fuliiru should bepossible. Considering that this book is likely to be the main and probably onlyreference on Fuliiru for many years to come, it is my sincere hope that a secondedition of Vol. 1 can be produced that addresses the many fixable problemsidentified here.


Bickmore, Lee. 2007. Cilungu phonology. Stanford, CA: CSLI Publications.

Bird, Charles. 1966. Aspects of Bambara syntax. Los Angeles: University ofCalifornia Ph.D. dissertation.

Clements, George N. & Kevin Ford. 1979. Kikuyu tone shift and its synchronicconsequences. Linguistic Inquiry 10. 179-210.

Downing, Laura J. 2001. Ungeneralizable minimality in Ndebele. Studies inAfrican Linguistics 30. 33-58.

Downing, Laura J. 2006. Canonical forms in prosodic morphology. Oxford: OxfordUniversity Press.

Downing, Laura J. 2011. Bantu tone. In Marc van Oostendorp, Colin J. Ewen,Elizabeth Hume & Keren Rice (eds.), The Blackwell companion to phonology.Cambridge, MA & Oxford: Blackwell.

Harjula, Lotta. 2004. The Ha language of Tanzania. Cologne: Rüdiger Köppe.

Hyman, Larry M. 2001. Privative tone in Bantu. In Shigeki Kaji (ed.),Cross-linguistic studies of tonal phenomena, 237-257. Tokyo: Institute for theStudy of Languages and Cultures.

Kisseberth, Charles & David Odden. 2003. Tone. In Derek Nurse & GérardPhilippson (eds.), The Bantu languages, 59-70. London: Routledge.

Park, Jae-Ick. 1997. Minimal word effects with special reference to Swahili.Bloomington, IN: Indiana University Ph.D. dissertation.

Paster, Mary & Yuni Kim. 2011. Downstep in Tiriki. Linguistic Discovery 9. 71-104.

Stevick, Earl. 1969. Tone in Bantu. International Journal of AmericanLinguistics 35. 330-341.


Michael R. Marlo is Assistant Professor of English and member of the Linguistics Program at the University of Missouri. His main research interests are in the study of tone and the description of underdocumented languages of Africa. He has been studying Kenyan varieties of the Luyia macrolanguage since 2000.

Page Updated: 17-Oct-2012