The Type Specimen (LB1) of Homo floresiensisDid Not Have Laron Syndrome

Dean Falk,1* Charles Hildebolt,2 Kirk Smith,2 William Jungers,3 Susan Larson,3 Michael Morwood,4

Thomas Sutikna,5 Jatmiko,5 E. Wahyu Saptomo,5 and Fred Prior6

1Department of Anthropology, Florida State University, Tallahassee, FL 32306-77722Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO 631103Department of Anatomical Sciences, Stony Brook University School of Medicine, Stony Brook, NY 117944University of Wollongong, School of Earth and Environmental Sciences, Wollongong, NSW, Australia 25225National Research and Development Centre for Archeology, Jakarta, Indonesia6Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO 63110

KEY WORDS Homo floresiensis; Laron Syndrome; microcephaly; pathology

ABSTRACT The type specimen (LB1) of Homo flore-siensis has been hypothesized to be a pathological humanafflicted with Laron Syndrome (LS), a type of primarygrowth hormone insensitivity (Hershkovitz et al.: Am JPhys Anthropol 134 [2007] 198–208). Comparing mea-surements, photographs and three-dimensional, com-puted-tomography reconstructions of LB1 with data anddiagnoses from the literature on LS, we critically evaluatenumerous skull and postcranial traits that Hershkovitzet al. identified as being shared by LB1 and patients withLS. The statements regarding most of these traits are newto the clinical literature and lack quantitative support.LB1 and patients with LS differ markedly in the size and

shape of the cranium; thickness and pneumatization ofcranial bones; morphology of the face, mandible, teeth,and chin; form of the shoulder, wrist, and pelvis; and gen-eral body proportions including relative foot size. Claimsthat patients with LS are similar to LB1 in displaying pro-tracted scapulae, short clavicles, low degrees of humeraltorsion, flaring ilia, and curved tibiae are not supportedby data or corroborating images. Some points of similarity(e.g., femoral neck-shaft angle, femoral bicondylar angle,and estimated stature) can be found in other hominins,and cannot be considered diagnostic. From our review andanalysis, we conclude that LB1 did not suffer from LS. AmJ Phys Anthropol 140:52–63, 2009. VVC 2009 Wiley-Liss, Inc.

Historically, announcements of new hominin speciesthat challenge conventional wisdom and dogma havebeen greeted with skepticism by vocal minorities of bothlaymen and scientists who question the new species’ au-thenticity and, instead, attribute the remains to unusual(or aberrant) apes or pathological humans (Gee, 2007).This happened with the discoveries of Neanderthals(Gruber, 1948; Trinkaus and Shipman, 1993; Drell, 2000;Regal, 2004), Homo (Pithecanthropus) erectus (Dubois,1896; Regal, 2004), and australopithecines (Dart, 1925;Findlay, 1972; Tobias, 1996), all of which were eventuallyrecognized as legitimate new taxa. We believe that it hashappened again in response to the 2004 announcement ofthe new species, Homo floresiensis (dubbed the ‘‘Hobbits’’)(Brown et al., 2004; Morwood et al., 2004, 2005), withsome workers proclaiming that the remains of the typespecimen, LB1, are those of a pathological Homo sapiens.Skeptics initially claimed that LB1 was nothing morethan a pathological person afflicted with microcephalyand disordered growth (Weber et al., 2005; Jacob et al.,2006; Martin et al., 2006a,b; Richards, 2006), althoughthis generic hypothesis has been repeatedly and thor-oughly refuted (Falk et al., 2005a,b, 2006, 2007a,b, 2008;Argue et al., 2006; Groves, 2007; Larson et al., 2007;Tocheri et al., 2007; Gordon et al., 2008; Jungers et al.,2008b; Lyras et al., 2008; Richards, 2006). Another recentand much more specific claim stated that LB1 was apathological Homo sapiens who was afflicted with a par-ticular form of primary growth hormone insensitivitycalled Laron Syndrome (LS) (Hershkovitz et al., 2007).The clinical phenotype associated with LS was first rec-

ognized in 1966 (Laron et al., 1966). Patients with LS

have normal to high serum levels of human growth hor-mone, but lack receptors for it in the liver, leading to adeficiency of the hormone somatomedin (IGF-I) thatcauses (if untreated) Laron-type dwarfism (Laron et al.,1966; Sarnat et al., 1988). In addition to short stature, 10other skeletal features have traditionally been identifiedas distinguishing patients with LS from nonaffected indi-viduals: 1) protruding forehead, 2) saddle nose, 3) shortface, 4) underdeveloped mandible, 5) small hands andfeet, 6) broken and discolored deciduous teeth and irregu-lar growth and crowding of the permanent dentition, 7)delicate long bones, 8) decreased bicondylar-biparietal(bic/bip) ratio, 9) small head circumference (HC), and 10)an upper/lower segment ratio above the norm for sex andage that denotes short limbs relative to trunk size (e.g.,Laron et al., 1968, 1979, 1991, 1992, 1993; Scharf andLaron, 1972; Konfino et al., 1975; Laron, 1995, 1999a,b,2004; Kornreich et al., 2002a,b).

Grant sponsor: National Geographic Society; Grant numbers:7769-04, 7897-05.

*Correspondence to: Dean Falk, Department of Anthropology,Florida State University, Tallahassee, FL 32306-7772, USA.E-mail: dfalk@fsu.edu

Received 15 July 2008; accepted 5 January 2009

DOI 10.1002/ajpa.21035Published online 17 March 2009 in Wiley InterScience

(www.interscience.wiley.com).

VVC 2009 WILEY-LISS, INC.

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 140:52–63 (2009)

One of the earliest descriptions of the syndrome notes‘‘From early childhood these children resembled eachother; the main feature was a small face and mandible,which gave a false impression of a large head. The dispro-portion between face and calvarium resulted in a saddlenose. . . . Aside from the short body, the hands and feetwere small (acromicria) . . . . teething was slow and theteeth discolored and broke easily . . . . growth of the perma-nent teeth was not only late but also irregular’’ (Laron etal., 1968, p 884–885). Further, ‘‘Aside from the retardationin skeletal maturation, X-rays of the skeleton revealedsmall, delicate long bones’’ (Laron et al., 1968, p 888). In1972, it was reiterated that ‘‘Acromicria and the slowdevelopment of the facial bones which cause protrusion ofthe frontal bones and saddle nose are characteristic ofhereditary-isolated growth-hormone deficiency’’ and thatpatients with LS ‘‘are very similar and at any age grouptheir bic/bip ratios are statistically significantly smallerthan those in the normal subjects’’ (Scharf and Laron,1972, p 93–94). Small HC (Laron et al., 1992) and shortlimbs relative to trunk size (Laron, 1999b) were also even-tually identified as typical for patients with LS.Over the years, these features continued to be listed as

typical clinical manifestations of LS; thus, in a 2004 retro-spective about his personal experience with LS from 1958to 2003, Laron reiterated all 10 features. In addition tothese traditionally, most-frequently-cited manifestationsof LS, Laron noted that the scant radiographic studiesthat have been conducted on the skulls of patients withLS revealed thin diploe, underdeveloped sinuses, delayedclosure of fontanelles and sutures, and a normal-sizedsella turcica (Laron, 2004).In 2007, Hershkovitz et al. provided a list of 33 ‘‘major

diagnostic criteria’’ of LS patients, 32 of which corre-sponded to traits described for LB1 by Brown et al. (2004;also cf. Morwood et al., 2005). The authors stated thatthese features demonstrated ‘‘morphological similarity

between individuals with LS and LB1’’ (Hershkovitzet al., 2007, p 2 and Table 1), indicating that the latter wasmost likely a modern human suffering from LS. Six of theten traditional diagnostic criteria for LS patients outlinedabove did not appear on the 2007 list of 33 criteria: pro-truding forehead, saddle nose, small hands and feet, bro-ken and discolored deciduous teeth and irregular growthand crowding of the permanent dentition, delicate longbones, and decreased bicondylar-biparietal ratio. Two ofthe 33 criteria contradicted earlier reports for LS: 1) pro-nounced humeral shaft thickness was listed as a diagnos-tic criterion for LS, although long bones of patients withLS have traditionally been described as delicate or thin,and 2) normal cranial thickness was listed as one of the33 diagnostic characteristics for LS, although the skulls ofLS patients have been described as having thin diploe andunderdeveloped sinuses (Laron, 2004). As discussedbelow, we have been unable to locate mention of many ofthe criteria that appear in LB1 and are now stated to bediagnostic for LS in the pre-2007 literature on LS. Thepurpose of this article is to address the scientific merits ofthe conclusion that LB1 and patients with LS share simi-lar diagnostic features, and that LB1 probably had LS.The suggestion that LB1 suffered from other pathologiessuch as cretinism (Obendorf et al., 2008) is addressed else-where (Jungers et al., 2009), as is the unsubstantiatedclaim (Henneberg and Schofield, 2008) that LB1 had adental filling in a mandibular molar (cf. Culotta, 2008 andthe link therein to Brown’s on-line rebuttal).

MATERIALS AND METHODS

Cranial and postcranial skeletal characteristics thatwere recently stated to be similar in LB1 and patientswith LS (Hershkovitz et al., 2007, Table 1) are evaluatedby comparing measurements, photographs, and radio-graphs obtained from the original Homo floresiensis

TABLE 1. Cranial measurements for LB1 and patients with LS

Cranial base ratiosA-P View: Bicondylar/Biparietal 3100a

LS patients, 8–20 years Males Females63.8 6 3 (n 5 5) 59.7 6 2.1 (n 5 9)

Normal controls 73.7 6 3.9 (n 5 22) 75.4 6 2.6 (n 5 18)LB1 (this study) (98.2/110.7) 3100 5 88.7

Lateral View: Basion-nasion/Cranial length 3100LS patients, 9–18 yearsb 48.9 (n 5 10)Normal controlsb 56.2 (n 5 10)LB1c 56.6Normal controlsc 55.2 (n � 3,400)

Head circumferenceLS patients, untreated (see Fig. 1)Normal adults (mean)d 58.0LB1 (cm, this study) 39.4

Frontal sinusesLS, untreated adultse

No sinus (n 5 6:3m, 3f)Diameter, cm \1.0 (n 5 3:1m, 2f)

Non-LS adults, 25 yearsf Males FemalesLength (cm) 1.74 6 0.52 (n 5 102) 1.61 6 0.58 (n 5 108)Width (cm) 2.80 6 0.71 (n 5 102) 2.64 6 0.67 (n 5 108)

LB1 (diameter, cm; this study) 2.09 (right side)

a Data (except for LB1) from Scharf and Laron, 1972:Table 1.b From Konfino et al., 1975: Table 1.c From Brown et al., 2004, Supplementary Table 1.d Hofman, 1984.e Kornreich et al., 2002a: Table 2.f Spaeth et al., 1997: Table 5; sample sizes from Table 1; all measurements right side.

53SICK-HOBBIT HYPOTHESIS

American Journal of Physical Anthropology

material [directly and from computed-tomography (CT)images] with those from the literature for patients withLS. The original CT data were obtained in Jakarta witha Siemens Emotion CT-scanner, and processed at Mal-linckrodt Institute of Radiology. See Falk et al. (2005a,2007a) for parameters that pertain to the CT imagesshown below.The following measurements were obtained from CT

images of LB1’s cranium:

Skull size and shape

1. Head circumference. We electronically measured HCin LB1 because it is used as a proxy for brain size in theLS literature. Although HC in patients with LS wasobtained by using a tape measure to determine the‘‘maximal perimeter around the glabella and the occiput’’(Laron et al., 1979, p 394), we modified the procedure toavoid inflating the estimate by including LB1’s supraor-bital torus (that patients with LS lack). The CT imagedata of LB1 were transferred using the digital imagingand communications in medicine (DICOM) format andloaded into the Mimics software package (Version 11.11Materialize, Ann Arbor, MI) where the skull was seg-mented (isolated from other parts of the image data) andrendered as a 3D model using the industry standardstereolithography format known as ‘‘stl.’’ The virtualskull model was loaded into Geomagic Studio software[Version 9 (SR 3) Geomagic, Research Triangle Park,NC] for visualization and quantification of the skull cir-cumference. A plane was defined using three points—one selected on the frontal bone directly above the supra-orbital torus and two selected at the maximum width ofthe skull. The intersection of this plane with the skull’sexterior defined the circumference contour. Data pointswere placed along the defined circumference and individ-ual, surface-contour length measurements were madeand summed to obtain the skull circumference.2. Skull asymmetry. The degree of facial asymmetry in

LB1 was investigated by using Geomagic Studio softwareto reconstruct right–right and left–left mirrored images ofLB1’s skull. The plane of section between the right andleft sides of the skull passed through the nasion, basion,and prosthion. These 3DCT mirrored images were com-pared visually with mirrored images obtained from 2Dphotographs of LB1’s face (Jacob et al., 2006).3. Cranial bone thickness. Thickness of LB1’s cranium

and diploe, and extent of pneumatization of the mastoidprocesses, paranasal sinuses, and cranial walls wereassessed visually using Mimics software.4. Areas of maximum cranial breadth.A. Bicondylar width. Because published measurements

of bicondylar/biparietal ratios are available for patientswith LS (Scharf and Laron, 1972), we obtained compara-ble measurements electronically for LB1. Geomagic Stu-dio software was used to measure the bicondylar widthof the skull. An interactive three dimensional (3D) viewwas used to measure the distance between the outerrims of LB1’s left and right mandibular fossae. To facili-tate this measurement, the mandible was segmentedand removed from the scene graph so that the mandibu-lar fossae were clearly visible during measurement. Datapoints were placed at the lateral margins of the fossae(per Scharf and Laron, 1972) and the linear distancebetween the points was measured electronically.B. Biparietal width. The CT image data for LB1 were

loaded into Plug 0n View 3D software (Version 3.1, Voxar

Limited, Boston, MA), which was used to measure thebiparietal width of the skull. A 3D view with linked cur-sors to axial, sagittal, and coronal views was used tolocate and identify the maximum width between the leftand right parietal bones (per Scharf and Laron, 1972).The linear distance between the identified points wasmeasured electronically.5. Frontal sinuses. Because of damage to LB1’s skull,

only the right frontal sinus is present. Its maximum diam-eter was measured electronically with Mimics software.6. Bifurcated root of premolar P3. Plug 0n View 3D

software was used to segment the right and left mandib-ular first premolars of LB1. The roots of these were visu-ally compared with the root of a developing mandibularright premolar in a cephalometric radiograph of a nor-mal child.

Postcranial bones

Hershkovitz et al. provided angular data on two ofeight postcranial features said to be similar in LS andLB1, femoral neck-shaft angle and femoral bicondylarangle. It proved difficult to assess rigorously the otherclaims made by Hershkovitz et al. with respect to limbproportions and the postcranium, because these wereunaccompanied by commensurate metric data for eitherLS patients and/or Homo floresiensis. We are able, how-ever, to provide data, observations, and images that arerelevant to their qualitative statements of similarity.

RESULTS WITH COMMENTARY

As detailed below, with respect to the 10 traditionaldiagnostic criteria for LS listed above, LB1 is not similarto patients with LS:1. Protruding forehead. LB1 lacks a protruding fore-

head that typically characterizes patients with LS (com-pare Fig. 1a,b). Instead, LB1 has a pronounced supraor-bital torus (rather than a ‘‘ridge’’) ‘‘that arches over theorbit and does not form a straight bar as in IndonesianH. erectus’’ (Brown et al., 2004, p 1057) (Fig. 1a,c).Patients with LS do not have supraorbital tori. Rather,the published lateral images of patients with LS (Rosen-bloom et al., 1999, Fig. 3; Laron, 2004, Fig. 4; Laron,1999a, Fig. 1; Laron, 1999b, Fig. 4; Vasil et al., 1994,Fig. 2b,c; Konfino et al., 1975, Fig. 1), which includevery few radiographs (Hershkovitz et al., 2007, Fig. 1;Vasil et al., 1994, Fig. 1d; Scharf and Laron, 1972, Fig.4), reveal a bulbous forehead (with frontal bossing) thatprotrudes forward relative to the face and differs mark-edly from LB1’s supraorbital torus (Fig. 1a,b).2. Saddle nose. It is difficult to assess the details of

the nasal bridge of LB1 because there is postmortemdamage to this region, but there is nothing to suggestthat LB1 shares the markedly depressed nasal bridge, or‘‘saddle nose,’’ that typifies patients with LS. In the lat-ter, the ‘‘prominent forehead, decreased vertical dimen-sion of face, hypoplastic nasal bridge, [and] small orbits’’(Kornreich et al., 2002a, p 499) cause the medial portionsof the supraorbital rims to be elongated relative to theirlateral portions, which give them a medially slantedappearance in frontal view (Scharf and Laron, 1972;Laron, 1999b, 2004) compared with the more roundedcontours of LB1 (compare Fig. 1c,d) and normal controls.This is consistent with the observation that, for patientswith LS, ‘‘the position of the nose root is steep and lowin relation to the forehead’’ (Konfino et al., 1975, p 200),

54 D. FALK ET AL.

American Journal of Physical Anthropology

which gives the nose a ‘‘pug’’ (saddle) shape in lateralview. The statement that patients with LS have roundedsupraorbital rims (Hershkovitz et al., 2007, Table 1) is,thus, not supported by the literature. Additionally, theorbits of LB1 are not unusually small (Brown et al.,2004, supplementary table 1), and there is no medialslant to the supraorbital rims.

3. Short face. Although reduced facial height has typi-cally been described for patients with LS (Hershkovitzet al., 2007, Table 1), this has always been in comparisonwith normal Homo sapiens. The facial height of LB1, onthe other hand, was described as reduced in comparisonwith Australopithecus, but not Homo sapiens (Brownet al., 2004, p 1057).4. Underdeveloped mandible. The mandible is not

underdeveloped in LB1, as is the case for patients withLS. The mandibles of patients with LS are ‘‘very small’’(Hershkovitz et al., 2007, p 3) compared (in multipledimensions) with normal Homo sapiens (Konfino et al.,1975), rather than ‘‘underdevelopment in the forwarddirection’’ (Hershkovitz et al., 2007, p 3). Despite lackinga chin, however, LB1’s mandible is not retrognathic, andit is relatively large compared with normal humans (Fig.1a), in keeping with its documented broad and thickenedramus and megadontia (Brown et al., 2004). Even in theimage of a patient with LS in their Erratum (Hershko-vitz et al., 2008), it is difficult to assess the area of themandible relevant to the development of the ‘‘chin.’’ Thisis because the image’s chin area lacks conspicuity. Thismay be attributable to this area’s being suboptimallyexposed to X-rays and/or to the image’s being subopti-mally processed/digitized. With Adobe1 Photoshop CS1

(Adobe Systems Incorporated, Seattle Washington, U.S.),we enhanced the area of the chin with image adjust-ments for shadow/highlight, with shadows set at 100%(Fig. 2a). In the enhanced image, the shape of chinappears to be the same as that of a normal human (asdepicted in lateral cephalogram images in the classic lit-erature, Broadbent and Golden, 1975) and is differentfrom the chin of LB1 (Fig. 2b). In addition, LB1 lacksthe diagnostic complex of external symphyseal featurestypical of modern people; this includes a distinctlyraised, everted, and inverted ‘‘T-shaped’’ structure that isassociated with flanking mental fossae (Schwartz andTattersall, 2000). A comparison of the lingual contours ofthe symphyseal region in LS and LB1 also reveals major

Fig. 1. Lateral (a) and frontal (c) views of LB1’s skull(adapted from Brown et al., 2004) compared with a typicalpatients with LS (b) from Hershkovitz et al., 2007; (d) adaptedfrom Scharf and Laron, 1972.

Fig. 2. (a) Figure from an erratum (Hershkovitz et al., 2008) purporting to show an underdeveloped chin in an individual suf-fering from Laron Syndrome. With Adobe1 Photoshop CS1 (Adobe Systems Incorporated, Seattle Washington, U.S.), we enhancedthe area of the chin with image adjustments for shadow/highlight, with shadows set at 100%. (b) A CT-based image of the craniumand mandible of LB1, midline section whitened (modified from Brown et al., 2004).

55SICK-HOBBIT HYPOTHESIS

American Journal of Physical Anthropology

differences. LB1 possesses superior and inferior trans-verse tori (not unlike those seen in some australopithe-cines such as LH4), whereas this feature is unknown inLS.5. Small hands and feet. The hands and feet of LS

individuals are known to be very small (acromicria;Laron, 2004). There is nothing in the partial hand ofLB1 to suggest that it is relatively small (Tocheri et al.,2007; Larson et al., 2008), and the reconstructed foot ofLB1 is relatively very long (Jungers et al., 2008a).6. Irregular growth and crowding of the permanent

dentition. It is not possible to assess the growth of LB1’spermanent dentition; yet in the original published photo-graphs of LB1 (Brown et al., 2004, Figs. 1 and 4) and Fig-ures 1, 2, and 3 of this article, there is no indication ofLB1’s permanent dentition that match the typical patternreported for patients with LS (Laron et al., 1968; Laron,1999a,b).7. Delicate long bones. The long bones of patients with

LS are typically described as being thin and delicate,although there is little quantitative published data forassessing this criterion of LS. Paradoxically, Hershkovitzet al. (2007) reported that ‘‘pronounced’’ thickness of thehumeral shaft relative to its length is characteristic of LSand LB1. In external dimensions, the humerus of LB1 isindeed robust (Morwood et al., 2005), but since no com-parative data were provided for LS, it is unclear how sim-ilar it may be to the humeri of patients with LS, and thearticles cited by Hershkovitz et al. (2007) supportingincreased bone thickness in LS deal with bone volume,including trabecular bone of the femoral neck and verte-brae, not thickness of the humerus. Subsequent to Hersh-kovitz et al. (2007), Kornreich et al. (2008) have docu-mented increased cortical thickness in the humeri ofpatients with LS. Since they defined a cortex as thickwhen the medullary cavity comprised one-third or less ofthe diameter of the shaft at the thickest cortical level, itis ambiguous whether they are actually identifyingincreased cortical thickness or medullary stenosis. In anyevent, CT scans indicate that the cortical bone of the LB1humerus is not especially thick, but rather is within thehuman range. Finally, Kornreich et al. (2008) alsoreported that the long bones of patients with LS are allproportionately small, in contrast to the robust externaldimensions of the LB1 humerus.8. Decreased bicondylar-biparietal ratio. The maxi-

mum cranial breadth in patients with LS is typicallylocated dorsally on the parietal bones, which causes adecreased bicondylar-biparietal ratio and gives skulls a

quasi-triangular shape in anterior view, as quantified bythe small bicondylar-biparietal ratios in patients with LScompared with normal controls (Scharf and Laron, 1972)(Fig. 1d, Table 1). This shape has been attributed tounderdevelopment of the base of the cranium comparedwith the neurocranium in LS (Scharf and Laron, 1972).In LB1, the shape of the skull differs dramatically fromthat of LS because the maximum breadth is in the‘‘inflated supramastoid region’’ (Brown et al., 2004,p 1056) near the bottom of the temporal bones (Fig. 1c).We obtained a bicondylar-biparietal ratio of 88.7 for LB1,which is not only 39% larger than that for persons withLS (males 5 63.8, females 5 59.7), but also larger thanthat for normal humans (males 5 73.7, females 5 75.4,Table 1; Fig. 1d, Fig. 3). Thus, the base of LB1’s skull iswell developed and, contrary to Hershkovitz et al.(2007), the area of maximum cranial breadth differs dra-matically between LB1 and persons with LS.We have also compiled data pertaining to the develop-

ment of the cranial base from a lateral perspective(basion-nasion/cranial length ratios) for patients withLS, normal controls, and LB1 (Table 1). Again, patientswith LS have cranial bases that are, at least, 11%smaller (48.9 versus 55.2) than normal controls, whileLB1 is only slightly, if at all, larger than normal controls(56.6 versus 55.2 or 56.2). In lateral view, the cranialvault of patients with LS appears relatively high (globu-lar) (Fig. 2a) because ‘‘the growing brain is forced todevelop upwards thus resulting in a protrusion of thefrontal bones and the bregma’’ (Scharf and Laron, 1972,p 95–96). In contrast, the LB1 cranial vault wasdescribed accurately as quite low (Brown et al., 2004)(Fig. 2b). Thus, cranial height/maximum cranial length(3100) for LB1 is 62.2, while it is 73.5 for 3,885 pooledHomo sapiens (see Brown et al., 2004, supplementarytable 1). (We have been unable to find published maxi-mum cranial heights for patients with LS.) Multiple pho-tographs in anterior–posterior views of patients with LS(Rosenbloom et al., 1999) also reveal a consistent skullshape that differs markedly from that of LB1, which isextremely brachycephalic (Fig. 1c,d): LB1’s maximumcranial breadth/maximum cranial length 3 100 5 79.0compared to 75.8 for 3,921 pooled Homo sapiens (seeBrown et al., 2004, supplementary table 1).9. Small head circumference. Although Hershkovitz

et al. (2007, p 3) noted that ‘‘the most outstanding featureof LB1 is the extremely small endocranial volume (410–380 cm3),’’ they used HC rather than cranial capacity asa surrogate for brain size. Some reviews of patients with

Fig. 3. Measurements obtained electronically for LB1’s head circumference (a), bicondylar width (b), and biparietal width (c).[Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

56 D. FALK ET AL.

American Journal of Physical Anthropology

LS, however, fail to list small HC as a main clinical fea-ture of LS (Laron et al., 1968, Fig. 4; Rosenfeld et al.,1994, Table 4) and published data suggest that, com-pared with LB1, brain size in patients with LS (to theextent that it can be inferred from HC) tends to be onlymoderately below the range for normal Homo sapiens(see Fig. 4). This is consistent with the observation that‘‘in effect the HC is below the normal size or in the lownormal ranges,’’ with some untreated LS patients havingnormal HCs (Laron et al., 1992, p 1260; Laron, 1999a,p 4402; 2004, p 1032). Similar curves for a total of sevenmales (some repeated in different publications) alsoshow HCs prior to treatment that range from nearly nor-mal to more than 2SD below the mean normal curves forboys (Laron et al., 1992, p 1260; Laron, 1995, p 1527;1999a, p 4402; 1999b, p 234). When plotted on the samegraph with untreated patients with LS, LB1’s HC (39.4cm) falls far below theirs, despite LB1’s being an adultand much older than these patients (see Fig. 4).Contrary to the condition seen in LB1, in an early study

of 22 patients with LS it was noted that a small face andmandible give ‘‘a false impression of a large head’’ (Laron etal., 1968, p 884). The observation that head size appearsrelatively large in patients with LS compared with normalpeople is echoed in many publications: ‘‘Craniofacial abnor-malities include subnormal HC and underdevelopment ofthe facial bones; nevertheless, head size is large relative tothe short stature’’ (Kornreich, 2002b, p 628). ‘‘We suggestthat in LS the disproportion between head and body sizesubjects the atlas and dens to increased stress, resulting inthickening and sclerosis at a young age’’ (ibid, p 630). ‘‘Thehead showing the typical appearance of a protruding fore-head, saddle nose, and sunset look seems large for thebody’’ (Laron, 2004, p 1032). In one review of LS (Rosenfeldet al., 1994, p 374), it was stated ‘‘the vertical dimension ofthe face is shortened; together with a normal head size andforehead length for stature, this may give an impression ofsmall facies (sic).’’ The appearance of a relatively largehead may contribute to an illusion of hydrocephalus(Rosenfeld et al., 1994; Rosenbloom et al., 1999). Thesenumerous observations indicate that head (brain) sizeappears large relative to stature in persons with LS com-pared with normal people, which is opposite the finding forbrain size relative to body size of LB1 (Falk et al., 2005a).10. Upper/lower segment ratio above the norm for sex

and age. Hershkovitz et al. described the limb propor-

tions of LB1 and LS as ‘‘abnormal.’’ The unusual limbproportions of LB1 refer to its high humerofemoral index(�86%) compared with modern people, including humanpygmies from Africa and Asia (Jungers, in press). Thisvalue recalls that of AL 288-1 (‘‘Lucy’’) at �85%. Theratio of reconstructed radius length to femur length inLB1 is also more similar to australopithecines than tohumans (Argue et al., 2006). Limb proportions are typi-cally assessed in LS via the ratio of upper to lower bodysegments (Laron, 1999b): sitting height is subtractedfrom stature to calculate lower segment length, and thena ratio is computed as sitting height divided by lowersegment length. This value tends to be high in LS incomparison with selected age-matched humans, but thecomparative data presented are sparse. At Age 18, onlythree LS data points are provided by Laron (1999b, Fig.6), and one of the three has a value less than two stand-ard deviations from the control sample. It is easy toderive this upper/lower ratio from the more traditionalindex of sitting height divided by stature, and there isgreat variation among modern human adults (Bogin andRios, 2003; Bogin and Beydoun, 2007). In fact, Mexican-American females who are just one standard deviationfrom their mean have upper to lower segment ratios of1.40, which encompasses the extreme range reported forLS 18-year-olds. Patients with LS have also been charac-terized as having a reduced arm span (Rosenfeld et al.,1994; Rosenbloom et al., 1999). Regardless, there is noway to assess interlimb proportions in LS from publisheddata, and there is no way to calculate sitting height (andits derivatives) in Homo floresiensis.

Other features said to link LB1 and patientswith LS (Hershkovitz et al., 2007)

1. Cranial facial asymmetry. Hershkovitz et al.(2007, p 2) stated that LB1’s facial and calvarial asymme-try exceeded clinical norms and were developmentallyabnormal, citing Jacob et al. (2006). The latter workersdivided a 2D photograph of LB1’s face along the midlineand used Photoshop (Adobe, San Jose, CA) to mirrorimage each half to create right–right and left–left compo-sites (Fig. 5, top). Because the two mirror-imaged compo-sites differed so much from each other, Jacob et al. con-cluded that the degree of craniofacial asymmetry in LB1exceeded clinical norms and that LB1 was, therefore,pathological. The representation of a 3D object with a2D image forces an observer to recreate complex 3Danatomy mentally. With 2D images, a slight error indetermining the midline of the face, including errorscaused by the angle at which the photograph was taken,may introduce bias. With 3D images, one can use 3Dlandmarks in performing mirror imaging tasks. Werepeated Jacob et al.’s experiment (Fig. 5, bottom row)by electronically bisecting and mirror imaging the twohalves of LB1’s entire skull along its midline, using3DCT data obtained from the original LB1 specimen(Falk et al., 2005a). The landmarks that we used werenasion, basion, and prosthion. Our results suggest anunremarkable and nonpathological degree of asymmetryin LB1’s face, contrary to Jacob et al. (2006). Weacknowledge that performing mirror imaging even with3D images is problematic; however, we think that per-forming mirror imaging with 3D data is far superior toperforming mirror imaging with 2D representations of a3D object.

Fig. 4. Head circumferences of untreated patients withLaron Syndrome at different ages, and for LB1. Notice thatLB1’s head circumference falls far below those of LS patients.

57SICK-HOBBIT HYPOTHESIS

American Journal of Physical Anthropology

This is not to say, however, that LB1’s skull is symmetri-cal. When viewed from the top, the left frontal and rightoccipital lobes of LB1’s virtual endocast (brain) are notice-ably wider and project further out from the brain thantheir counterparts (Falk et al., 2005a, Fig. 1). This left-frontal and right-occipital petalia pattern is the reverse ofthe pattern seen in most human brains (and skulls) and isstatistically correlated with left-handedness (LeMay,1992)—especially if the pattern is marked (Bear et al.,1986), as it is in LB1. The asymmetry of LB1’s skull wasassessed with 3D geometric morphometrics (Baab andMcNulty, 2008), and it was concluded that its degree ofleft–right asymmetry is within the ranges observed fornormal humans and apes (and comparable to that seen inother fossil hominins that have never been characterizedas pathologically asymmetrical).2. Cranial bone thickness. According to Hershkovitz

et al. (2007, Table 1), LB1 and patients with LS bothhave normal cranial bone thickness. Neither statementis supported by the literature. Adult patients with LSare characterized as having ‘‘thin cranial vaults’’ and(for 11 patients) ‘‘the diploe of the calvarial bones wasthin in all the patients’’ (Kornreich et al., 2002a, p 500).In a review of 60 patients with LS, Laron (2004, p 1034)noted: ‘‘The diploe of the skull is very thin and thesinuses are underdeveloped.’’ These observations areconsistent with an earlier one that ‘‘retardation in skele-tal maturation also affects the membranous neurocra-nium’’ (Laron, 1999a, p 4400). Rather than cranial bonethickness that is normal for humans, ‘‘medially, later-ally, and basally, the cranial vault bone is thick’’ in LB1and, further, ‘‘LB1 is relatively thickened in areas ofpneumatization in the lateral cranial base’’ (Brown et al.,

2004, p 1057). These observations are supported bynumerous measurements from the original material(ibid: supplementary table 1) as well as visual inspectionof LB1’s CT images (see Fig. 6). LB1’s thick cranial vaultbones and expanded diploe are within the recordedrange of variation for Homo erectus (Brown, 1994;Brown et al., 2004) and should not be misconstrued aspathological.3. Pneumatization of the skull. Hershkovitz et al. (2007)

stated that ‘‘nonpneumatized (acellular)’’ skulls typifypatients with LS and, further, suggested that this featureoffers an opportunity for hypothesis testing by examiningradiological data from LB1 (Hershkovitz et al., 2007, p 3):‘‘Unfortunately, no radiographs of LB1’s skull are as yetavailable and therefore appreciation of the extent of pneu-matization in the LB1 skull is impossible. Nonpneuma-tized (acellular) mastoid process [characterizes] LS.’’LB1’s CT images show that the cranial diploe that arewidely noted as ‘‘thin’’ in patients with LS (but excludedfrom Hershkovitz et al., 2007, Table 1) are, instead, pneu-matized in parts of LB1’s braincase, as are LB1’s mastoidprocesses (Fig. 6a,c).4. Frontal sinuses. Table 1 of Hershkovitz et al. lists the

frontal sinuses as being ‘‘absent/undersized’’ in patientswith LS, which is verified by measurements from the liter-ature (Kornreich et al., 2002a). For the frontal sinus onthe right side of LB1 (Fig. 6b), we obtained a diameter of20.9 mm, which is larger than those reported for patientswith LS (Kornreich et al., 2002a) and sorts LB1 with nor-mal adults (Table 1). This is remarkable given that LB1’scranium is much smaller than those of either normal oradult patients with LS (see Fig. 4); therefore, the state-ment that ‘‘cranial features shared by both LS and LB1’’

Fig. 5. Asymmetry in LB1’s skull. Normal views on left; mirrored right–right and left–left images are in the center and right,respectively. Top row is from a 2D photograph of LB1’s face (modified from Jacob et al., 2006); bottom: from 3DCT data from LB1’sentire skull (this study). The mandible was not included in our reconstruction because it was not clear that its anatomical relationto the maxilla was optimally positioned for performing a mirror imaging task. [Color figure can be viewed in the online issue whichis available at www.interscience.wiley.com.]

58 D. FALK ET AL.

American Journal of Physical Anthropology

includes ‘‘absence of or undersized frontal sinuses’’(Hershkovitz et al., 2007, p 3) is incorrect.5. Bifurcated root of premolar P3. This feature is pres-

ent in LB1 (Fig. 7a–d) and ‘‘recorded for Australopithe-cus and early Homo, and some Indonesian H. erectus’’mandibles (Brown et al., 2004, p 1058). Hershkovitz et al.(2007, p 4–5) claimed that in LS ‘‘the second [meaningP4] mandibular premolars have double roots (Fig. 5)’’(See Fig. 7e). The second mandibular premolar in Figure7e appears, however, to be a normally developing P4whose root is in the process of forming and whose apexhas not formed. Teeth such as these are common in radi-ographs of children’s teeth (e.g., Fig. 7f) and can be seenin oral radiology text books (White and Pharoah, 2000).6. Clavicular size and shape. Larson et al. (2007)

reported that the clavicle of LB1 is short, and concludedthat this implied a protracted scapula. Hershkovitz et al.(2007) stated, without supporting metrics, that theclavicles of individuals with LS were also short, but theirown figure (see their Fig. 6) failed to reveal a protractedscapula. They also described the clavicle as being‘‘shallow,’’ which we take to mean a relatively straightclavicular shaft. If this is correct (again, there is no sup-porting documentation), then this does not characterizethe clavicle of LB1 (Larson et al., 2007, 2008).7. Humeral torsion. Hershkovitz described humeral

torsion in both LS and LB1 as ‘‘limited,’’ or relativelyreduced in comparison with most modern people.Reduced humeral torsion in LB1 (and other fossil homi-nins) has indeed been documented (Larson et al., 2007),but Hershkovitz et al. (2007) presented no data to bol-ster their claim about this condition in LS. Kornreichet al. (2008, p 159) noted that many patients with LSdisplayed unexplained limited elbow extensibility, andsince a few of the patients in their sample displayed arotated ulna when the humerus was in a presumed neu-tral position, they speculated that the limited elbow exten-sibility in LS could be related to an abnormally highdegree of humeral head retroversion (5 reduced torsion).They did not explain how elbow extension could be influ-enced by the degree of humeral torsion and offered no datato support their supposition indicating that while humeralretroversion could be measured on CT, it was not done forethical reasons. We note, however, that studies on humeral

retroversion in living human subjects are routinely doneusing simple plain radiographs (e.g., Kronberg et al., 1990;Osbahr et al., 2002; Reagan et al., 2002) or ultrasound(e.g., Ito et al., 1995; Whiteley et al., 2006), and do not nec-essarily require CT.8. Iliac flare. Hershkovitz et al. (2007) described lat-

eral flaring of the iliac blade as ‘‘marked‘‘ in both LB1and LS despite commenting that they had no clear ideawhat Brown et al. (2004) meant specifically in their ini-tial description of marked lateral iliac flare in LB1. Theradiographs they offered to document iliac flare qualita-tively in LS are less than compelling in our opinion.Here we show what is meant by pronounced lateral iliacflare in LB1 in comparison with a LS patient, and sub-mit that there is no apparent similarity between the two(see Fig. 8). To the contrary, the iliac flare seen in LB1recalls that seen is australopithecines (Lovejoy, 2005)and Homo erectus (Simpson et al., 2008).9. Femoral neck-shaft angle. Femoral neck-shaft angle

was described as 130 degrees in LB1 by Brown et al.(2004) and slightly less at 128 degrees by Jungers et al.(2008b). Hershkovitz et al. (2007) reported a range of118–134 degrees in LS. Of note, 128–130 degrees is nearthe mean value for many modern human groups (Bello yRodriguez, 1909; Grine et al., 1995, Table 4), and wellwithin the range of various groups of hunter/gatherer/foragers (including North African and European Meso-lithic peoples, Jomon, Australian aborigines and Khoe-san; also see Trinkaus, 1993). This ‘‘similarity’’ betweenLS and LB1 is therefore unremarkable.10. Femoral bicondylar angle. Brown et al. (2004)

reported a value of 14 degrees for the femoral bicondylarangle of LB1. Hershkovitz et al. provided a range of 10–16 degrees for LS individuals. The value for LB1 is quitesimilar to some australopithecines (9–15 degrees) andcharacterizes some early modern Homo (Tardieu andTrinkaus, 1994), although it would be unusually high formost truly modern humans. It follows that the valuereported for LB1 has no special affinity with anglesdescribed for patients with LS.11. Tibial long axis. The long axis of the tibia of LB1

(Brown et al., 2004) and that of LS are said to be‘‘curved’’ (Hershkovitz et al., 2007); however, the smalleradult LB8 tibia appears to have a very straight diaphy-

Fig. 6. Cranial Bone Thickness. Contrary to characteristics that are typical for patients with LS, CT scans of LB1’s craniumreveal thick diploe (a, arrow), a right frontal sinus with a diameter of 2.1 cm (b, arrow), and highly pneumatized mastoid processes(c, arrows). Note also, the general overall thickness of the calvaria.

59SICK-HOBBIT HYPOTHESIS

American Journal of Physical Anthropology

sis (Jungers et al., 2008b). In view of the variability seenin the axial curvature of human tibiae (Jungers et al.,2008b), and lacking relevant comparative data, this pro-posed similarity carries no diagnostic weight.

DISCUSSION

As documented above, LB1 is different from patientswith LS with respect to the 10 features that have tradi-tionally been associated with LS; nor is there evidence tosupport Hershkovitz et al.’s conclusion that LB1 closely

and/or uniquely resembles patients with LS in the follow-ing additional features: cranial facial asymmetry, cranialbone thickness, degree of pneumatization of the skull, sizeof frontal sinuses, bifurcated root of premolar P3, clavicu-lar size and shape, humeral torsion, iliac flare, femoralneck-shaft angle, femoral bicondylar angle, and curvatureof the long axis of the tibia. As we have documented, LB1and patients with LS also differ in their relative brainsizes, the extent to which they are prognathic, the mor-phologies of their supraorbital rims and supraorbitalridges, and the forms of their mandibular symphyses.

Fig. 8. (a) Radiograph of a22-year-old man with Laron Syn-drome (modified from Kornreichet al., 2008); (b) left os coxae ofLB1 articulated with a cast ofthe sacrum of AL 288-1 (Austral-opithecus afarensis); (c) femalepelvis; (d) male pelvis; (e)australopithecine pelvis (c–eadapted from Lovejoy, 2005). Thesimilarity between b and e isnoteworthy. [Color figure can beviewed in the online issue whichis available at www.interscience.wiley.com.]

Fig. 7. Mandibular right(a, b) and left (c, d) first premo-lars of LB1; (b and d) the premo-lars (which are fully formed)have been segmented from themandible; (e) from Figure 5 ofHershkovitz et al., 2007 (arrowadded for emphasis). The captionfor this figure reads: ‘‘Mandibleof a child with LS. Note the dou-ble roots of the second premo-lar’’; (f) lateral cephalometric ra-diograph of a normal child. Theright and left second premolarsare overlapping. As is normal,the roots of the premolars arestill undergoing developmentand root apices have not formed,which gives the illusion of a dou-ble-root. [Color figure can beviewed in the online issue whichis available at www.interscience.wiley.com.]

60 D. FALK ET AL.

American Journal of Physical Anthropology

Although Hershkovitz et al. (2007, p 2) stated thattheir analyses were based on ‘‘data of 64 patients fol-lowed for the past 45 years by one of the authors (ZL)from infancy to adulthood, [and] direct observation of ra-diographs and CT images (including three-dimensionalrendering method) of 15 adult (age 21–68 years) patients(seven males, eight females) with LS,’’ publications thatdescribe the skeletal morphology for LS are extremely‘‘scant’’ (Laron, 2004, p 1034) and contain little in theway of measurements and quantitative analyses that arerelevant to the debate over Homo floresiensis. Further,there are few published radiographic (let alone CT)images of untreated patients with LS, and those thatexist are of poor quality and have been republished overmany years. The recent article by Kornreich et al. (2008)is a welcome new addition in this regard.Our analysis demonstrates that, except for short stat-

ure and nondiagnostic angles of the femur, LB1 looksnothing like patients with LS, as confirmed by the illus-trations in this article, measurements from the originalLB1 specimen, the relevant data that are available inthe LS literature, and our analyses of ‘‘diagnostic crite-ria’’ that are listed in Hershkovitz et al.’s Table 1 (2007).As discussed above, the degree of asymmetry seen inLB1’s face does not appear to be outside the normalrange for humans, and, as noted, the petalia pattern ofLB1’s skull and endocast suggest left-handedness whenseen in normal humans. Further, there is no evidence inthe LS literature or in Hershkovitz et al. (2007) to sup-port the latter’s statement that patients with LS haveendocranial surfaces (endocasts) that resemble LB10s.Nor is the inference that LB1 had LS strengthened by

the observation that ‘‘one should not expect completecranial morphological similarity between our group of[LS] patients and the single LB1 skull’’ because of thenumerous mutations and ‘‘large cranial morphologicalvariability’’ involved in LS (Hershkovitz et al., 2007,p 7–8). Despite all of the variation in LS, Hershkovitz et al.have not described nor illustrated even one patient withLS who looks anything like LB1.Finally, Hershkovitz et al. also offered an observation

regarding LS that is related to the hypothesis that LB1was afflicted with microcephaly. Although they 1) com-pared LB1 favorably to the Rampasasa pygmy populationreferred to by Jacob et al. (2006) and interpreted the‘‘scant anthropological reports available on the skulls ofthe prehistoric island population’’ as suggesting ‘‘a smallcranium with projected face and posteriorly sloped frontalbone,’’ and 2) viewed the microcephalic hypothesis as ten-able, Hershkovitz et al.’s analysis excluded LS patientsthat had microcephaly (2007, p 7–8). Falk et al. (2007a)measured, compared, and illustrated brain shape featuresin virtual endocasts from nine microcephalic and ten nor-mal humans and provided a mathematical classificationfunction that completely separates the two groups. Whenthe function was applied to LB1, it sorted with the bigger-brained normal humans rather than the microcephalics(although other features distinguish LB1’s endocast fromthose of Homo sapiens). This study strongly suggests thatthe shape of Hobbit’s endocast is the antithesis of a micro-cephalic’s (Falk et al., 2007a,b, 2008).

CONCLUSION

Insofar as we can tell, LB1, who lived to be a matureadult female, did not suffer from LS and was not patholog-ical. It is fascinating that the discovery of Homo floresien-

sis has received from a vocal minority of scientists thesame kind of skepticism that befell the initial discoveriesof Neanderthals, Homo erectus, and australopithecines.The doubts regarding LB1 may be associated with a beliefthat any hominin with such a small brain that lived asrecently as 18,000 years ago must have been pathological.If so, this is in spite of equally small-brained, tool-produc-ing hominins being widely known from earlier parts of thefossil record (Falk et al., 2000; Falk and Clarke, 2007). Aswe have documented, however, neither the LS nor micro-cephalic hypothesis regarding LB1 withstands close scien-tific scrutiny.We understand the cognitive dissonance that the discov-

ery of Homo floresiensis has created in some scientificcircles, and we encourage efforts to frame testable, alter-native hypotheses to account for these surprising homi-nins. We submit that ‘‘pathology,’’ however, is not a scien-tific explanation unless a differential diagnosis is madespecific, plausible, and testable. The LS hypothesis, whichwe emphatically reject as the correct diagnosis for LB1,was explicit and had testable components, and we believethat this is how science ultimately progresses.

ACKNOWLEDGMENTS

The authors thank the Wenner-Gren Foundation forsupporting efforts to conserve the fossils. The lateralcephalometric radiograph (Fig. 7f) was provided byDebra Dixon (Southern Illinois University School of Den-tal Medicine). Luci Betti-Nash assisted with some of theartwork. They thank Yoel Rak and two anonymousreviewers for constructive suggestions about this article.

LITERATURE CITED

Argue D, Donlon D, Groves C, Wright R. 2006. Homo floresiensis:microcephalic, pygmoid, Australopithecus, or Homo? J HumEvol 51:360–374.

Baab KL, McNulty KP. 2008. Scaling and asymmetry in fossilhominins: the status of the LB1 cranium based on 3D mor-phometric analyses. J Hum Evol; doi:10.1016/j.jhevol. 2008.08.011.

Bear DM, Schiff D, Saver J, Greenberg M, Freeman R. 1986.Quantitative analysis of cerebral asymmetry; fronto-occipitalcorrelation, sexual dimorphism and association with handed-ness. Arch Neurol 43:598–603.

Bello y Rodriguez S. 1909. Le femur et le tibia chez l’homme etchez les anthropoıdes (Variations suivant le sexe, l’age et larace). Paris: G. Jacques.

Bogin B, Beydoun N. 2007. The relationship of sitting heightratio to body mass index and fatness in the United States,1988–1994. Hum Ecol Special Issue 15:1–8.

Bogin B, Rios L. 2003. Rapid morphological change in livinghumans: implications for modern human origins. Comp Bio-chem Physiol (A) 136:71–84.

Broadbent BB, Golden WH. 1975. Bolton standards of dentofa-cial developmental growth. St. Louis: CV. Mosby.

Brown P. 1994. Cranial vault thickness in Asian Homo erectusand Homo sapiens. Courier Forschungs-Institut Senckenberg171:33–46.

Brown P, Sutikna T, Morwood MJ, Soejono RP, Jatmiko, Sap-tomo EW, Awe Due R. 2004. A new small-bodied homininfrom the Late Pleistocene of Flores, Indonesia. Nature 431:1055–1061.

Culotta E. 2008. Tempest in a hobbit tooth. Science NOW April24, 2008. Available at http://sciencenow.sciencemag.org/cgi/content/full/2008/424/2.

Dart RA. 1925. Australopithecus africanus: the man-ape ofSouth Africa. Nature 115:195–199.

61SICK-HOBBIT HYPOTHESIS

American Journal of Physical Anthropology

Drell JRR. 2000. Neanderthals: a history of interpretation.Oxford J Arch 19:1–24.

Dubois E. 1896. On Pithecanthropus erectus: a transitional formbetween man and the apes. J Anthropol Inst Great BritainIreland 25:240–255.

Falk D, Clarke R. 2007. New reconstruction of the taung endocast:a brief communication. Am J Phys Anthropol 134:529–534.

Falk D, Hildebolt C, Smith K, Morwood MJ, Sutikna T, BrownP, Jatmiko, Saptomo EW, Brunsden B, Prior F. 2005a. Thebrain of LB1, Homo floresiensis. Science 308:242–245.

Falk D, Hildebolt C, Smith K, Morwood MJ, Sutikna T, Jatmiko,Saptomo EW, Brunsden B, Prior F. 2005b. Response to com-ment on ‘‘The brain of LB1, Homo floresiensis.’’ Science 310:236c.

Falk D, Hildebolt C, Smith K, Morwood MJ, Sutikna T, Jatmiko,Saptomo EW, Brunsden B, Prior F. 2006. Response to commenton ‘‘The brain of LB1, Homo floresiensis.’’ Science 312:999c.Available at www.sciencemag.org/cgi/content/full/312/5776/999c(full text).

Falk D, Hildebolt C, Smith K, Morwood MJ, Sutikna T, Jatmiko,Saptomo EW, Imhof H, Seidler H, Prior F. 2007a. Brain shapein human microcephalics and Homo floresiensis. Proc NatlAcad Sci USA 104:2513–2518.

Falk D, Hildebolt CG, Smith KE, Prior F. 2007b. LB1’s virtualendocast: implications for hominin brain evolution. In: IndriatiE, editor. Proceedings of the International Seminar on South-east Asian Paleoanthropology: Recent Advances on SoutheastAsian Paleoanthropology and Archaeology. Laboratory ofBioanthropology and Paleoanthropology, Faculty of MedicineGadjah Mada University, Yogyakarta, Indonesia, p 37–46.

Falk D, Hildebolt C, Smith K, Morwood MJ, Sutikna T, Jatmiko,Saptomo EW, Prior F. LB1’s virtual endocast, microcephaly,and hominin brain evolution. J Hum Evol, in press. doi:10.1016/j.jhevol.2008.10.008.

Falk D, Redmond JC, Guyer J, Conroy GC, Recheis W, WeberGW, Seidler H. 2000. Early hominid brain evolution: a newlook at old endocasts. J Hum Evol 38:695–717.

Findlay GH. 1972. Dr. Robert Broom, F. R. S. Palaeontologistand Physician/1866–1951. Cape Town: AA Balkema.

Gee H. 2007. In a hole in the ground [Book Review of ‘‘The dis-covery of the Hobbit: the Scientific Breakthrough thatChanged the Face of Human History’’ (by MJ Morwood and Pvan Oosterzee).] Nature 446:979–980.

Gordon AD, Nevell L, Wood B. 2008. The Homo floresiensis cra-nium (LB1): size, scaling, and early Homo affinities. Proc NatlAcad Sci USA 105:4650–4655.

Grine FE, Jungers WL, Tobias PV, Pearson OM. 1995. Fossil HomoFemur from Berg Aukas. Am J Phys Anthropol 97:151–185.

Groves C. 2007. The Homo floresiensis controversy. HAYATI JBiosci 14:123–126.

Gruber JW. 1948. The Neanderthal controversy: nineteenth-cen-tury version. Sci Monthly 67:436–439.

Henneberg M, Schofield J. 2008. The Hobbit trap, money, fame,science and the discovery of a ‘new species’. Adelaide, SouthAustralia: Wakefield Press.

Hershkovitz I, Kornreich L, Laron Z. 2007. Comparative skele-tal features between Homo floresiensis and patients with pri-mary growth hormone insensitivity (Laron Syndrome). Am JPhys Anthropol 134:198–208.

Hershkovitz I, Kornreich L, Laron Z. 2008. ERRATUM: I.Hershkovitz, L. Kornreich, Z. Laron (2007). Comparative skel-etal features between Homo floresiensis and patients with pri-mary growth hormone insensitivity (Laron Syndrome). Am JPhys Anthropol 134:198–208. Am J Phys Anthropol 136:373.

Hofman MA. 1984. A biometric analysis of brain size in micren-cephalics. J Neurol 231:87–93.

Ito M, Eto M, Maeda K, Rabbi ME, Iwasaki K. 1995. Ultrasono-graphic measurement of humeral torsion. J Shoulder ElbowSurg 4:157–161.

Jacob T, Indriati E, Soejono RP, Hsu K, Frayer DW, Eckhardt RB,Kuperavage AJ, Thorne A, Henneberg M. 2006. Pygmoid Aus-tralomelanesian Homo sapiens skeletal remains from LiangBua, Flores: population affinities and pathological abnormal-ities. Proc Natl Acad Sci USA 103:13421–13426.

Jungers WL. Interlimb proportions in humans and fossil homi-nins: variability and scaling. In: Grine FE, Leakey RE, Flea-gle JG, editors. The first humans: origins of the genus Homo.Springer: Dordrecht, in press.

Jungers WL, Falk D, Hildebolt C, Smith K, Prior F, Tocheri M,Orr CM, Burnett SE, Larson SG, Djubiantono T, MorwoodMJ. 2009. The hobbits (Homo floresiensis) were not cretins.Am J Phys Anthropol Suppl 48:244.

Jungers WL, Harcourt-Smith WEH, Larson SG, Morwood MJ,Djubiantono T. 2008a. Hobbit bipedalism: functional anatomyof the foot of Homo floresiensis. Am J Phys Anthropol Suppl46:127.

Jungers WL, Larson SG, Harcourt-Smith W, Morwood MJ, Suti-kna T, Rokus Due Awe, Djubiantono T. 2008b. Descriptions ofthe lower limb skeleton of Homo floresiensis. J Hum Evol; doi:10.1016/j.jhevol. 2008.08.014.

Konfino R, Pertzelan A, Laron Z. 1975. Cephalometric measure-ments of familial dwarfism and high plasma immunoreactivegrowth hormone. Am J Orthod 66:196–201.

Kornreich L, Horev G, Schwarz M, Karmazyn B, Laron Z.2002a. Craniofacial and brain abnormalities in Laron syn-drome (primary growth hormone insensitivity). Eur J Endo-crinol 146:499–503.

Kornreich L, Horev G, Schwarz M, Karmazyn B, Laron Z.2002b. Laron Syndrome abnormalities: spinal stenosis, osodontoideum, degenerative changes of the atlanto-odontoidjoint, and small oropharyne. Am J Neuroradiol 23:625–631.

Kornreich L, Konen O, Schwarz M, Siegel Y, Horev G, Hershko-vitz I, Laron Z. 2008. Abnormalities of the axial and proximalappendicular skeleton in adults with Laron syndrome (growthhormone insensitivity). Skeletal Radiol 37:153–160.

Kronberg M, Brostrom LA, Soderlund V. 1990. Retroversion ofthe humeral head in the normal shoulder and its relationshipto the normal range of motion. Clin Orthop 253:113–117.

Laron Z. 1995. Prismatic cases: Laron Syndrome (primarygrowth hormone resistance) from patient to laboratory topatient. J Clin Endocrinol Metab 80:1526–1531.

Laron Z. 1999a. The essential role of IGF-I: lessons from thelong-term study and treatment of children and adults withLaron Syndrome. J Clin Endocrinol Metab 84:4397–4404.

Laron Z. 1999b. Natural history of the classical form of primarygrowth hormone (GH) resistance (Laron Syndrome). J PediatrEndocrinol Metab 12:231–249.

Laron Z. 2004. Laron syndrome (primary growth hormone re-sistance or insensitivity). The personal experience 1958–2003.J Clin Endocrinol Metab 89:1031–1044.

Laron Z, Anin S, Klipper-Aurbach Y, Klinger B. 1992. Effects ofinsulin-like growth factor on linear growth, head circumfer-ence, and body fat in patients with Laron-type dwarfism. Lan-cet 339:1258–1261.

Laron Z, Klinger B, Grunebaum M. 1991. Laron type dwarfism.Special feature—Picture of the month. Am J Dis Child 145:473–474.

Laron Z, Pertzelan A, Karp M. 1968. Pituitary dwarfism withhigh serum levels of growth hormone. Isr J Med Sci 4:883–894.

Laron Z, Pertzelan A, Karp M, Keret R, Eshet R, Silbergeld A.1993 Laron Syndrome: a unique model of IGF-I deficiency.History, clinical and biochemical characteristics. In: Laron Z,Parks JS, editors. Lessons from Laron Syndrome (LS) 1966–1992. Pediatr Adolesc Endocrinol. Basel, Karger 24:3–23.

Laron Z, Pertzelan A, Mannheimer S. 1966. Genetic pituitarydwarfism with high serum concentration of growth hormone-anew inborn error of metabolism? Isr J Med Sci 2:152–155.

Laron Z, Roitman A, Kauli R. 1979. Effect of human growthhormone theapy on head circumference in children with hypo-pituitarism. Clin Endocrinol 10:393–399.

Larson SG, Jungers WL, Morwood MJ, Sutikna T, Jatmiko,Saptomo EW, Due RA, Djubiantono T. 2007. Homo floresiensisand the evolution of the hominin shoulder. J Hum Evol 53:718–731.

Larson SG, Jungers WL, Tocheri MW, Orr CM, Morwood MJ,Sutikna T, Due RA, Djubianto T. 2008. Descriptions of the

62 D. FALK ET AL.

American Journal of Physical Anthropology

upper limb skeleton of Homo floresiensis. J Hum Evol; doi: 10.1016/j.jhevol. 2008.06.007.

LeMay M. 1992. Left-right dissymmetry, handedness. Am JNeuroradiol 13:493–504.

Lovejoy CO. 2005. The natural history of human gait and pos-ture, Part 1: Spine and pelvis. Gait Posture 21:95–112.

Lyras GS, Dermitzakis MD, Van der Geer AAE, Van derGeer SB, DeVos J. 2008. The origin of Homo floresiensisand its relation to evolutionary processes under isolation.Anthropol Sci J-Stage (www.jstage.jst.go.jp.); doi:10.1537/ase.080411.

Martin RD, MacLarnon AM, Phillips JL, Dussebieux L, Wil-liams PR, Dobyns WV. 2006a. Comment on ‘‘The brain of LB1,Homo floresiensis.’’ Science 312:999b.

Martin RD, MacLarnon AM, Phillips JL, Dobyns WB. 2006b.Flores hominid: new species or microcephalic dwarf? Anat RecA 288:1123–1145.

Morwood MJ, Brown P, Jatikmo, Sutikna T, Saptomo EW, West-away KE, Due RA, Roberts RG, Maeda T, Wasisto S, Djubian-tono T. 2005. Further evidence for small-bodied homininsfrom the Late Pleistocene of Flores, Indonesia. Nature 437:1012–1017.

Morwood MJ, Soejono RP, Roberts RG, Sutikna T, Turney CSM,Westaway KE, Rink WJ, Zhao J-x, van den Bergh GD, DueRA, Hobbs DR, Moore MW, Bird MI, Fifield K. 2004. Archae-ology and age of a new hominin species from Flores in easternIndonesia. Nature 431:1087–1091.

Obendorf PJ, Oxnard CE, Kefford BJ. 2008. Are the smallhuman-like fossils found on Flores human endemic cretins?Proc R Soc B 275:1287–1296.

Osbahr DC, Cannon DL, Peer KP. 2002. Retroversion of thehumerus in the throwing shoulder of college baseball pitchers.Am J Sports Med 30:347–353.

Reagan KM, Meister K, Horodyski MB, Werner DW, CarruthersC, Wilk K. 2002. Humeral retroversion and its relationship toglenohumeral rotation in the shoulder of college baseball play-ers. Am J Sports Med 30:354–360.

Regal B. 2004. Human evolution: a guide to the debates. SantaBarbara, CA: ABC-CLIO.

Richards GD. 2006. Genetic, physiologic and ecogeographic fac-tors contributing to variation in Homo sapiens: Homo flore-siensis reconsidered. J Ev Biol 19:1744–1767.

Rosenbloom AL, Guevara-Aguirre J, Rosenfeld RG, Francke U.1999. Growth hormone receptor deficiency in Ecuador. J ClinEndocrinol Metab 84:4436–4443.

Rosenfeld RG, Rosenbloom AL, Guevara-Aguirre J. 1994.Growth hormone (GH) insensitivity due to primary GH recep-tor deficiency. Endocrinol Rev 15:369–390.

Sarnat H, Kaplan I, Pertzelan A, Laron A. 1988. Comparison ofdental findings in patients with isolated growth hormone defi-ciency treated with human growth hormone (hGH) and inuntreated patients with Laron-type dwarfism. Oral Surg 66:581–586.

Scharf A, Laron Z. 1972. Skull changes in pituitary dwarfismand the syndrome of familial dwarfism with high plasma im-munoreactive growth hormone. Horm Metab Res 4:93–97.

Schwartz JH, Tattersall I. 2000. The human chin revisited:what is it and who has it? J Hum Evol 38:367–409.

Simpson SW, Quade J, Levin NE, Butler R, Dupont-Nivet G,Everett M, Semaw S. 2008. A female Homo erectus pelvisfrom Gona, Ethiopia. Science 322:1089–1092.

Spaeth J, Krugelstein U, Schlundorff G. 1997. The paranasalsinuses in CT-imaging: development from birth to age 25. IntJ Pediatr Otorhinolaryngol 39:25–40.

Tardieu C, Trinkaus E. 1994. Early ontogeny of the human fem-oral bicondylar angle. Am J Phys Anthropol 95:183–195.

Tobias PV. 1996. Premature discoveries in science with especialreference to Australopithecus and Homo habilis. Proc Am PhilSoc 140:49–64.

Tocheri MW, Orr CM, Larson SG, Sutikna T, Jatmiko, SaptomoEW, Due RA, Djubiantono T, Morwood MJ, Jungers WL.2007. The primitive wrist of Homo floresiensis and its implica-tions for hominin evolution. Science 317:1743–1745.

Trinkaus E. 1993. Femoral neck-shaft angles of the Qafzeh-skhul early modern humans, and activity levels among imma-ture Near Eastern Middle Paleolithic hominids. J Hum Evol25:393–416.

Trinkaus E, Shipman P. 1993. The Neandertals. Changing theimage of mankind. NewYork: Alfred A. Knopf.

Vasil M, Baxova A, Kozlowski K. 1994. Radiographic abnormal-ities in Laron dwarfism. Pediatr Radiol 24:260–262.

Weber J, Czarnetzki A, Pusch CM. 2005. Comment on ‘‘Thebrain of LB1, Homo floresiensis.’’ Science 310:236b.

White SC, Pharoah MJ. 2000. Oral radiology: principles andinterpretation. St. Louis, MO: Mosby.

Whiteley R, Ginn K, Nicholson L, Adams R. 2006. Indirectultrasound measurement of humeral torsion in adolescentbaseball players and non-athletic adults: reliability and signif-icance. J Sci Med Sport 9:310–318.

63SICK-HOBBIT HYPOTHESIS

American Journal of Physical Anthropology