Energy and Buildings 110 (2016) 71–78

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

Short communication

A taxonomy of vernacular architecture

An addendum to “Ancient vernacular architecture:
Characteristics categorization and energy performance evaluation”
(Zhai and Previtali, 2010)
Jonathan M. Previtali, Zhiqiang (John) Zhai ∗
Department of Civil, Environmental and Architectural Engineering, University of Colorado, UCB 428, ECOT 441, Boulder, CO 80309, USA

a r t i c l e i n f o a b s t r a c t

Article history: The paper is an addendum to the paper entitled “Ancient vernacular architecture: characteristics catego-
Received 5 October 2015 rization and energy” that was published at “Energy and Buildings” in 2010 [1]. This addendum provides
Accepted 6 October 2015 an explicit list of 114 codes representing vernacular architectural regions across the whole globe defined
Available online 8 October 2015
by climate, culture and continent. Codes are intended to support work in analyzing and memorializing
vernacular building characteristics that otherwise may be lost. This addendum will be an important and
Keywords:
useful resource for analyzing and studying vernacular architecture.
Taxonomy
© 2015 Elsevier B.V. All rights reserved.
Vernacular architecture

1. Introduction The taxonomy documented here, in a great detail, provides a

Shaped by a similar adaptive process as biological evolution,
vernacular architecture holds intrinsic qualities to support heating,
cooling and the use of natural materials that in many cases have
enabled ancient cultures to live in ways that are less energy and
resource intensive than modern societies. There is value in iden-
tifying, documenting and categorizing vernacular techniques for
the sake of archiving this knowledge built up over centuries and
potentially applying these techniques to today’s structures.
Zhai and Previtali [1] presented the overall approach to
categorizing distinct vernacular regions, and evaluating energy per-
formance of ancient vernacular technologies and homes through
systematic building energy simulations. The simulation results
reveal that integrating traditions (both materials and techniques)
found in ancient vernacular architecture as a measure to improve
building energy performance is a worthwhile endeavor and a sci-
entific guidance can help enhance the performance. The study also
indicates that, since many vernacular dwells exist in the world,
it is challenging (but desired) to package vernacular architecture
traditions and quantitative design knowledge to modern building
designers. In-depth analysis of vernacular architectures in a par-
ticular location is thus necessary to provide location and culture
specific design solutions.
∗ Corresponding author.
E-mail address: john.zhai@colorado.edu (Z. Zhai).
methodology and system to categorize distinct vernacular regions
and what is believed by the author is a global, comprehensive list
that will assist the mentioned performance analysis, evaluation and
optimization.
2. Categorizing vernacular regions
Since vernacular traditions are shaped strongly by culture,
weather and geographic location, this taxonomy classifies the
world into distinct vernacular architecture regions each with a
unique combination of these three traits. It is important to note
that GPS mapping boundaries have not been created for the tax-
onomy and are not recommended so as to allow for flexibility in
categorizing structures on boundaries or in islands of climate or
climate within larger vernacular architectural regions.
2.1. Mapping climate zones
Köppen (1846–1940) developed a classification system that
is generally accepted as the most accurate method of mapping
world climates. It includes 14 zones separated by temperature and
humidity. However, the Köppen climate zone map was found to
be considerably complex for the purpose of this taxonomy and a
simplified map based on the Köppen Climate Classification System
developed by Richard de Dear at Macquarie University in Australia
was chosen [2] (Fig. 1).

http://dx.doi.org/10.1016/j.enbuild.2015.10.015
0378-7788/© 2015 Elsevier B.V. All rights reserved.
72 J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78

Fig. 1. A simplified mapping of world climates [2].

2.2. Mapping cultural heritage Table 1

Number of vernacular regions per climate zone.

Given the migration of peoples and the temporal nature of Climate zone Number of vernacular regions
national borders, the question arises of how to map cultural Continental subarctic (CS) 10
heritage. A relatively accurate method generally accepted by Tundra ice cap (TIC) 6
anthropologists is the tracing of language families. A language fam- Desert (D) 13
ily is a grouping of languages based on linguistic similarities. While High altitude (HA) 11
Humid mid latitude (HML) 8
religions, geographic locations, regional languages and even skin
Mediterranean (M) 11
color change within a related people, basic language traits such as Semi desert (SD) 15
syntax, phonetics, and semantics are strong indicators of a shared Tropical savanna (TS) 13
cultural heritage. West coast marine (WCM) 9
Wet equatorial (WE) 11
A map of language families was drawn from the Evolution of the
Humid subtropical (HS) 7
Human Language Project (aka The Tower of Babel: http://starling. Total 114
rinet.ru/intrab.php?lan=en), a compilation of linguistic research
started by Sergei Anatolyevich Starostin in 1997 and continuously
augmented by other linguistics researchers as new data becomes

Fig. 2. World map of language families.

 J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78 73

Table 2
Taxonomy nomenclature.

Climate zone Language family Continent

CS Continental sub arctic AA Afro-Asiatic AN Antarctica

D Desert AL Altaic AF Africa
HA High Altitude AM Amerind AS Asia
HML Humid Mid Latitude AUS Australian AU Australian
HS Humid Subtropical AUC Austric EU Europe
M Mediterranean CK Chukchi-Kamchatkan NA North America
SD Semi Desert DC Dene-Caucasian SA South America
TIC Tundra Ice Cap DR Dravidian OC Oceania
TS Tropical Savana EA Eskimo-Aleut
WCM West Coast Marine IE Indo-European
WE Wet Equatorial IP Indo-Pacific
KA Kartvelian
KH Khoisan
KJA Korean, Japanese, Ainu, Gilyak
NK Niger-Kordofanian
NS Nilo-Saharan
UR Uralic

Table 3
Codes for mapped vernacular regions.

Climate Language Africa (AF) Asia (AS) Australian Europe North America South America Oceania (OC)
Zone Family (AU) (EU) (NA) (SA)

CS AL CS-AL-AS
CS AM CS-AM-NA
CS CK CS-CK-AS
CS DC CS-DC-AS CS-DC-NA
CS EA CS-EA-NA
CS IE CS-IE-EU
CS UR CS-UR-AS CS-UR-EU
D AA D-AA-AF
D AL D-AL-AS
D AM D-AM-NA D-AM-SA
D DC D-DC-AS D-DC-NA
D DR D-DR-AS
D IE D-IE-AS
D KH D-KH-AF
D NK D-NK-AF
D NS D-NS-AF D-NS-AS
HA AA HA-AA-AF
HA AL HA-AL-AS
HA AM HA-AM-NA HA-AM-SA
HA AUS HA-AUC-OC
HA DC HA-DC-AS HA-DC-NA
HA IE HA-IE-AS HA-IE-EU
HA IP HA-IP-OC
HML AL HML-AL-AS
HML AM HML-AM-NA
HML DC HML-DC-AS HML-DC-NA
HML IE HML-IE-EU
HML KJA HML-KJA-AS
HML UR HML-UR-AS HML-UR-EU
HS AM HS-AM-NA HS-AM-SA
HS DC HS-DC-AS HS-DC-NA
HS KJA HS-KJA-AS
HS NK HS-NK-AF
M AA M-AA-AF M-AA-AS
M AL M-AL-AS
M AM M-AM-NA M-AM-SA
M AUC M-AUS-AU
M DC M-DC-NA
M IE M-IE-AS M-IE-EU
M KA M-KA-EU
M KH M-KH-AF
SD AA SD-AA-AF SD-AA-AS
SD AL SD-AL-AS
SD AM SD-AM-NA SD-AM-SA
SD AUC SD-AUS-AU
SD DC SD-DC-AS SD-DC-EU
SD IE SD-IE-AS SD-IE-EU
SD KA SD-KA-EU
SD NK SD-NK-AF
SD NS SD-NS-AF
SD UR SD-UR-AS
74 J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78

Table 3 (Continued)

Climate Language Africa (AF) Asia (AS) Australian Europe North America South America Oceania (OC)
Zone Family (AU) (EU) (NA) (SA)

TIC EA TIC-EA-EU TIC-EA-NA

TIC IE TIC-IE-EU
TIC UR TIC-UR-AS TIC-UR-EU
TS AA TS-AA-AF
TS AM TS-AM-NA TS-AM-SA
TS AUS TS-AUC-AF TS-AUC-AS TS-AUC-OC
TS AUC TS-AUS-AU
TS DC TS-DC-AS
TS DR TS-DR-AS
TS IE TS-IE-AS
TS KH TS-KH-AF
TS NK TS-NK-AF
TS NS TS-NS-AF
WCM AM WCM-AM-NA WCM-AM-SA
WCM AUS WCM-AUC-AU (1)
WCM AUC WCM-AUS-AU (1)
WCM DC WCM-DC-EU (1) WCM-DC-NA (1)
WCM EA WCM-EA-NA (1)
WCM IE WCM-IE-EU (1)
WCM IP WCM-IP-AU (1)
WE AM WE-AM-NA (1)
WE AUS WE-AUC-AF (1) WE-AUC-OC (1)
WE AUC WE-AUS-AU (1)
WE DC WE-DC-AS (1)
WE DR WE-DR-AS (1)
WE IE WE-IE-AS (1)
WE IP WE-IP-OC (1)
WE NK WE-NK-AF (1)
WE NS WE-NS-AF (1)

available. To keep the number of vernacular regions close to 100, it
was decided to use the first level of language families for all groups
other than Eurasiatic which was broken into second level language
families. The resulting map includes 17 language families with a
good distribution across continents (Fig. 2).
between continents such as mountains, oceans, etc., create barriers
to migration causing cultures to remain in place and develop new
vernacular traditions. Their use also facilitates quick identification
of a region.

2.4. Combining maps to identify unique vernacular regions

2.3. Mapping continents
Combining the de Dear climate map and the map derived from
Continental boundaries were used as the final characteristic the Evolution of the Human Language Project and the continental
in developing regions with the idea that geographical barriers map, resulted in 114 vernacular regions across 11 climate zones

Fig. 3. Continental subarctic (CS).

 J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78 75

Fig. 4. Tundra ice cap (TIC).

that contained a unique combination of language family, climate
and continent (Table 1). The process used to combine the maps
was visual. A mapping of each climate zone was simply laid over
the map of language families to reveal vernacular regions. Clearly
this process did not result in highly well defined boundaries for
each vernacular region, but it was deemed acceptable because each
region is intended to highlight an area where one would likely find
a unique example of vernacular architecture.
3. Naming the vernacular regions
stands for Semi Desert–Amerind–North America. Table 2 lists the
taxonomy nomenclature for all climate zones, language families
and continents that are identified and studied.
4. Codes for mapped vernacular regions
Table 3 presents the entire code lists for mapped vernacular
regions.

For categorization convenience, the study developed a nomen-
clature for the vernacular regions, by combining acronyms for each
climate zone, language family and continent into a hyphenated
code. The sequence was chosen for no particular purpose to be Cli-
mate Zone – Language Family – Continent; for instance, SD–AM–NA
5. Climate zone maps showing vernacular regions and
codes
Figs. 3–13 are the climate zone maps showing the proposed
vernacular regions and codes.

Fig. 5. Humid mid latitude (HML).

76 J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78

Fig. 6. West coast marine (WCM).

Fig. 7. High altitude (HA).

Fig. 8. Mediterranean (M).

J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78 77

Fig. 9. Semi desert (SD).

Fig. 10. Desert (D).

Fig. 11. Humid subtropical (HS).

78 J.M. Previtali, Z. Zhai / Energy and Buildings 110 (2016) 71–78

Fig. 12. Tropical savanna (TS).

Fig. 13. Wet equatorial (WE).

References [2] R. de Dear, Developing an Adaptive Model of Thermal Comfort and Preference,
ASHRAE RP – 884, 1997.
[1] J. Zhai, J.M. Previtali, Ancient vernacular architecture: characteristics
categorization and energy performance evaluation, Energy Build. 42 (3) (2010)
357–365.