See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/273489027

The Book of Trees: Visualizing Branches of Knowledge. -- By Manuel Lima. *

Design for Information: An Introduction to the Histories, Theories, and Best
Practices Behind Effective I...

Article in Systematic Biology · February 2014

DOI: 10.1093/sysbio/syu092

CITATIONS READS

3 6,617

1 author:

David A Morrison
Uppsala University
247 PUBLICATIONS 7,258 CITATIONS

SEE PROFILE

All content following this page was uploaded by David A Morrison on 05 September 2016.

The user has requested enhancement of the downloaded file.

 Book Reviews

Syst. Biol. 64(2):363–365, 2015

© The Author(s) 2014. Published by Oxford University Press, on behalf of the Society of Systematic Biologists. All rights reserved.
For Permissions, please email: [email protected]
DOI:10.1093/sysbio/syu092
Advance Access publication November 21, 2014

The Book of Trees: Visualizing Branches of when referring to the different “branches” of human
Knowledge. By Manuel Lima. New York: Princeton knowledge (e.g., branches of science). For example,
Architectural Press, 2014. 208 pp. ISBN 978-1-61689-218-0 Joachim of Fiore used the tree as a metaphor for historical
$29.95. £18.99 (hardback). relationships in his Liber Figurarum of 1202 (Hestmark
2000), a book whose exquisite prints could easily have
Design for Information: An Introduction to been included above under “art.” In his book Arbor
the Histories, Theories, and Best Practices Behind Scientiae Venerabilis et Cælitus of 1295, Ramón Llull used

Downloaded from http://sysbio.oxfordjournals.org/ by David Morrison on February 19, 2015

Effective Information Visualizations. By Isabel the tree to illustrate the growth and inter-relationships
Meirelles. Beverly MA: Rockport Publishers, 2013. 224 of knowledge more generally (Gontier 2011; Kutschera
pp. ISBN 978-1-59253-806-5 $40.00 (paperback). 2011).
It is with this role of trees as illustrations of information
Trees can be many things: objects, art, symbols, or that the two review books concern themselves. They
information. thus represent the latest manifestations of a very long
As objects, trees act as homes and shelter, they provide tradition involving visualizations of human knowledge.
food and oxygen, and they bind soil to hold topography The tree is probably the most ubiquitous and long-
in place. You can read about this in any biology textbook. lasting of our visual metaphors, illustrating the relations
Trees have also had a long tradition in the visual arts. between objects as well as the relations between
To me, perhaps the most interesting books about trees as concepts. In the modern world the Tree of Knowledge
art are those by Fowles and Horvat (1979) and by Shyam has been greatly generalized, so that trees are now
et al. (2006). The former is a disquisition by novelist John both visual and mathematical representations of the
Fowles on the connection between the natural world and relationships among pieces of information. There is thus
human creativity, accompanied by moody photographs much that is new for these two authors to discuss,
of trees taken by Frank Horvat. The latter book is a series because computers and new algorithmic models have
of hand-lithographed prints of tribal art images from produced an array of new methods and designs.
three Gond people of central India (Bhajju Shyam, Durga The book by Manuel Lima (The Book of Trees) focuses
Bai, and Ramsingh Urveti). (And yes, the land of the on trees exclusively, although some of them you may
Gond is Gondwanaland, which was the source of our not have recognized as trees. The book is arranged by
name for the southern land masses.) type of tree: Figurative trees, Vertical trees, Horizontal
The most famous use of trees as symbols is the Tree trees, Multidirectional trees, Radial trees, Hyperbolic
of Life, which recurs in many cultures throughout the trees, Rectangular treemaps, Voronoi treemaps, Circular
world, and which you can read about in Cook (1974). treemaps, Sunbursts, and Icicle trees. A tree is defined as
It often appears as a World Tree, which supports the representing hierarchically structured information, and
heavens, thereby connecting the heavens, the human therefore any such representation can be called a “tree,”
world, and (through its roots) the underworld. This including things that look more like maps and Venn
motif has appeared in specific forms in many cultures, diagrams than like traditional trees.
including Assyrian, Akkadian, Sumerian, Babylonian, Each chapter is arranged chronologically, which
Egyptian, Ancient Greek, Nordic (Norse), Celtic, Olmec, acknowledges the historical milieu noted above
Aztec, Mayan, Buddhist, Tibetan, Hindu, and Siberian, (covering more than 500 years). This means that the
among others. The Biblical Tree of Life, on the other information being represented is not arranged by
hand, was actually the lignum vitae (Tree of Eternal context, and thus conceptual themes recur throughout
Life) not the arbor vitae (Tree of Life). It was explicitly the book rather than being consolidated. For example,
contrasted with the lignum scientiae boni et mali (Tree of phylogenetic trees have historically been drawn
Knowledge of Good and Evil), from which Adam and as figurative, vertical, horizontal, multidirectional,
Eve ate the forbidden fruit, and were thereby ejected radial, or hyperbolic (the latter being restricted
from Eden. to interactive trees), and so phylogenetic trees are
This Biblical imagery was later co-opted as the illustrated throughout the book. Moreover, some trees
arbor scientiae (Tree of Knowledge), starting with the could easily fit into more than one chapter, such as the
Porphyrian tree in the third-century AD (although there horizontal trees on pp. 98–101, which could as easily be
are no extant copies from that time). That is, knowledge seen as multidirectional.
can often be arranged like the branches of a tree; and This means that the book provides only a visual
indeed, that metaphor has come down to us today overview of trees, rather than providing some sort of

363

[17:56 3/2/2015 Sysbio-syu092.tex] Page: 363 363–367

 364 SYSTEMATIC BIOLOGY VOL. 64

critical commentary and intellectual review. The book (timelines and flows), Spatial structures (maps), Spatio-
simply starts with one type of tree diagram and ends temporal structures, and Textual structures. Meirelles
with another type. This is the book’s biggest weakness— thus paints a much broader picture of data visualization
it focuses on the visual characteristics and historical and information design than does Lima, with trees as
circumstances, rather than on the theoretical aspects. I merely one possible iconography—in terms of coverage
strongly felt the lack of a critical overview of what trees her book is broader but shallower.
can and cannot do in terms of representing information. Meirelles also has a much stronger pedagogic
A tree can be a clever way of displaying data, but philosophy than does Lima. Representing
that does not mean that it is necessarily a clear way multidimensional information in two dimensions
of displaying it, and this distinction is insufficiently is not trivial, and Meirelles’ goal is to provide a
emphasized in the book. combined discussion of the technical requirements and
It is thus important to recognize that a beautiful- the design aspects of such visualizations. She insists
looking tree does not necessarily represent information that “understanding the constraints and capabilities
accurately. As but one recent example, D’Efilippo of cognition and visual perception is essential”
and Ball (2013) produce a classic Tree of Life drawn (p. 9), because her emphasis is on design rather than

Downloaded from http://sysbio.oxfordjournals.org/ by David Morrison on February 19, 2015

as a real tree (as inaugurated by Haeckel 1866). information. Her writing thus focuses much more on
Unfortunately, quite a number of the taxonomic labels the theory, while still insisting that it is successful
are misplaced, and we are therefore treated to some practice that is the ultimate goal.
rather surprising pieces of alleged phylogenetic and Phylogenetic diagrams are few in her book, and all of
systematic information. them are gathered in the chapter on trees. Surprisingly,
Nevertheless, Lima’s book does present us with none of the examples used appear in the book by Lima.
a pictorial buffet of the sheer variety of Trees of As an aside, I am not sure that either author actually
Knowledge, both in terms of what they can look like understands phylogenetic trees. For example, one of
and what information they can convey. The oldest Lima’s labels starts with: “One of the first phenetic
images are the most beautiful, of course, as modern diagrams, also known as ‘cladograms,’ produced by
ones have become more stylized, adhering to Edward numerical methods” (p. 103). (This figure was allegedly
Tufte’s dictum that information is most clearly displayed published in something called the “Oxford Journal
by using the minimum amount of ink possible. I think of Systematic Biology”, which you and I know as
that Tufte is right in general, but I do miss the old, more Systematic Zoology; this is far from the only bibliographic
discursive, style. error in his book.) Furthermore, neither author shows
The treemaps are probably the tree images that are any interest in anthropology in either of its venerable
most unfamiliar to systematists and phylogeneticists, guises as linguistics (language studies) or stemmatology
who have traditionally favored lines to display (manuscript studies), both of which have long used trees
informational connections (i.e., the “node-link” tree to represent genealogies sensu lato.
layout), rather than using nested areas (i.e., space filling) Interestingly, Meirelles shows that treemap methods
to represent hierarchical information. Nevertheless, this were used in systematics long before the invention of
is what treemaps do—each branch of the tree is given an computers (Lima notes only that they first appeared in
area, which is then tiled with smaller areas representing cartography in 1845). Indeed, we can produce a treemap
sub-branches. The main advantage of using a map as a if we simply cut horizontal slices out of a vertical tree,
representation is that the size and color of the areas can as shown on p. 28 of Design for Information, which
be used to represent other information about each tree reproduces Maximilian Fürbringer’s (1888) tree of bird
leaf. relationships. On the left is the side view of the tree, and
This idea has, on occasion, been adopted in biology. on the right are three slices through the tree branches
For example, taxonomic hierarchies are sometimes (as viewed from above). The latter produces a circular
represented using a treemap, such as in the web treemap, which is admittedly a less efficient use of the
database BioNames (which displays the taxonomic visualization space compared to a rectangular one.
groups recognized by the Index to Organism Names In spite of the chapter on relational structures,
database), and the Natural Science Museum of Barcelona phylogenetic networks do not explicitly appear in
(which allows interactive access to the database records Meirelles book, which I personally consider to be a major,
via a taxonomic hierarchy). It has also been used if forgivable, oversight. The book’s networks are strictly
to display the gene ontology associated with gene of the type that connects observed nodes via observed
expression data from microarray studies (Baehrecke et al. links, rather than connecting observed (leaf) nodes via
2004) as well as other ’omics data. It has even been inferred (internal) nodes and inferred links. That is,
suggested that treemaps could be used to represent none of them use the network to represent temporal
phylogenetic trees (Arvelakis et al. 2005). relationships, let along evolutionary ones.
Trees and treemaps also appear in the book by Nevertheless, phylogenetic networks of a sort do
Isabel Meirelles (Design for Information), although this appear in the book. A network is not a nested hierarchy,
book covers much more than that. This time the book but instead involves a collection of overlapping sets.
is arranged by subject: Hierarchical structures (trees), This can be represented as a Venn diagram, for example,
Relational structures (networks), Temporal structures but not as a treemap. This form of visualization has

[17:56 3/2/2015 Sysbio-syu092.tex] Page: 364 363–367

 2015 BOOK REVIEWS 365

also been a longstanding suggestion in phylogenetics. Baehrecke E.H., Dang N., Babaria K., Shneiderman B. 2004.
Two of these appear in the book. One is Goldfuss’ Visualization and analysis of microarray and gene ontology data
(1817) set of nested egg-shaped sets, expressing his ideas with treemaps. BMC Bioinform. 5:84.
Cook R. 1974. The Tree of Life: image for the cosmos. London: Thames
about affinity relationships, with one set overlapping & Hudson.
several of the others, representing a nonnested series of D’Efilippo V., Ball J. 2013. The infographic history of the world. London:
relationships. The other is Swainson’s (1837) quinarian HarperCollins.
figure of nonoverlapping sets, expressing his ideas Fowles J., Horvat F. 1979. The tree. London: Aurum Press.
Fürbringer M. 1888. Untersuchungen zur Morphologie und Systematik
about multidimensional affinity relationships. Oddly, der Vögel. T.J. Amsterdam: van Holkema.
both figures are in the Hierarchy chapter not the Goldfuss G.A. 1817. Über de Entwicklungsstufen des Thieres.
Relational one, presumably reflecting the nested part of Nürnberg: Leonhard Schrag.
their information rather than the concomitant reticulate Gontier N. 2011. Depicting the Tree of Life: the philosophical and
information. historical roots of evolutionary tree diagrams. Evol. Educ. Outreach
4:515–538.
Both books, by Lima and Meirelles, are beautifully Haeckel E. 1866. Generelle Morphologie der Organismen. Berlin: G.
produced in full color, and the images dominate rather Reimer.
than the text. If you do not mind the limited range of Hestmark G. 2000. Temptations of the tree. Nature 408:911.

Downloaded from http://sysbio.oxfordjournals.org/ by David Morrison on February 19, 2015

iconography, then these books can be viewed as being Kutschera U. 2011. From the scala naturae to the symbiogenetic and
about art just as much as information. That is, they dynamic tree of life. Biol. Direct 6:33.
Shyam B., Bai D., Urveti R. 2006. The night life of trees. Chennai, India:
show us infographics in the original sense, before people Tara Books.
started putting cute cartoons all over them to distract Swainson W. 1837. On the natural history and classification of birds.
attention from the actual information. I recommend both London: Longman et al.
books as a feast for sore scientific eyes.

REFERENCES
David A. Morrison, Systematic Biology, Evolutionary Biology
Arvelakis A., Reczko M., Stamatakis A., Symeonidis A., Tollis I.G. 2005. Centre, Norbyvägen 18D, 752 36 Uppsala, Sweden; E-mail:
Using treemaps to visualize phylogenetic trees. LNCS 3745:283–293. [email protected]

Syst. Biol. 64(2):365–367, 2015

© The Author(s) 2014. Published by Oxford University Press, on behalf of the Society of Systematic Biologists. All rights reserved.
For Permissions, please email: [email protected]
DOI:10.1093/sysbio/syu127
Advance Access publication December 22, 2014

Homology, Genes, and Evolutionary Innovation. should use character identity as the basis for the
Günter P. Wagner. Princeton: Princeton University homology of morphological characters. For readers of
Press, 2014. xv+478 pp. ISBN: 978-0-691-15646-0 $60.00 Systematic Biology, the book is also a reminder that every
£41.95 (hardback). morphological character used in a phylogenetic analysis
is a hypothesis of homology, and that great care is needed
Günter Wagner has produced a magisterial treatment when deciding whether morphological characters in
of one of the truly fundamental concepts in evolutionary different organisms are likely to be homologs.
biology—homology. Homology is also one of the most Wagner’s wide-ranging book is divided into two parts
controversial topics in biology, and tends to mean with 13 chapters (plus Preface and Introduction). Part
different things to different types of biologists. In I covers Concepts and Mechanisms and consists of:
molecular biology, homology is sometimes used as a Chapter 1 The Intellectual Challenge of Morphological
synonym for “similarity,” for example stating that two Evolution: A Case for Variational Structuralism,
genes in different organisms are “77% homologous” Chapter 2 A Conceptual Roadmap to Homology,
when their DNA sequences show 77% similarity. To Chapter 3 A Genetic Theory of Homology, Chapter
evolutionary biologists, this is like stating that someone 4 Evolutionary Novelties: The Origin of Homologs,
is “42% married,” because we tend to think of homology Chapter 5 Developmental Mechanisms for Evolutionary
as either/or statements—two structures are either Novelties, Chapter 6 The Genetics of Evolutionary
homologous or not. This is just one example of how Novelties, and Chapter 7 The Long Shadow of
different biologists use the term homology in very Metaphysics on Research Programs. Having laid the
different ways because it is hard to identify how two conceptual foundation, the author then, in Part II,
parts can be homologous in spite of differences in form, surveys examples that serve to illustrate how the
function, and underlying genetic mechanisms. concepts developed in Part I can be used in empirical
Günter Wagner has thought long and hard about research programs. Part II thus covers Paradigms and
homology in relation to character identity, and in his Research Programs, including: Chapter 8 Cell Types
new book he goes into great detail about why we and Their Origins, Chapter 9 Skin and a Few of Its

[17:56 3/2/2015 Sysbio-syu092.tex] Page: 365 363–367

View publication stats