Détails

Mis à jour : 6 novembre 2014

Cette partie du site est actuellement en construction...

This section of the site is currently under construction...

 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Détails

Mis à jour : 6 novembre 2014

Cette partie du site est actuellement en construction...

This section of the site is currently under construction...

Détails

Mis à jour : 9 novembre 2014

Cette partie du site est actuellement en construction...

This section of the site is currently under construction...

 A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z

A

***Agnatic (Uterine) Component – It is the largest component of the agnatic (uterine) chains network.

***Agnatic (Uterine) Weight – It is the number of individuals whose linear agnatic (uterine) ascendant of a given generation is known, as a percentage of all individuals for whom the at least one (agnatic or uterine) ascendant of that generation is known.

***Agnatic (Uterine) Net Weight - It is the number of individuals for whom only the agnatic (uterine) ascendant of a given generation is known, as a percentage of individuals for whom at least one (agnatic or uterine) ascendant of that generation is known.

***Alliance Matrix – It is the matrix of a matrimonial alliance network. It is a contingency (or cross) weighted table that indicates the number of spouses "exchanged" between a given set of classes/groups of individuals. As the classes are listed in the first top-line and first left-column, the descending diagonal of the matrix represent the endogamic (intra-group) marriages.

B

***Bicomponent - A bicomponent (or bi-connected component) is a graph where two distinct Paths can connect  any two vertices to each other. Thus, a bicomponent contains no cut-point whose elimination would cut it into two disconnected components. Consequently, any two vertices in a bicomponent form part of a cycle (Grange & Houseman, 2010; Hamberger & Daillant, 2008).

C

Chain - A chain is a (sub)graph whose vertices and arcs form a single path. Using this notion presuppose a socio-centered view of kinship ties.

Consanguine Chain - It is a kinship chain that contains neither parental triads nor (marriage) edges.

Cycle - A cycle is a path where the first and the last vertex are identical. Using this notion presuppose an ego-centered view of kinship ties.

Circuit - A circuit is a (sub)graph whose vertices and arcs form a single cycle. Using this notion presuppose a socio-centered view of kinship ties.

Circuit Census - See matrimonial census.

***Circuit Composition Table – It is a symmetrical cross table that indicates which matrimonial circuit type is the product of the intersection of two other types of circuit (for an example of a circuit composition table, limited to the third canonical degree, see Hamberger & Daillant, 2009).

***Circuit Inclusion – It is a specific kind of intersection between two circuits, where one contains all the other’s individuals.

***Circuit Intersection – It is a relation between two circuits, which share one or more marriage ties.

***Circuit Intersection Matrix – It is the matrix of a circuit intersection network. It indicates the occurrences of each circuit type, as well as the number of marriages involved in a circuit intersection.

***Circuit Intersection Network – It is the network of all the circuit intersections in a given kinship network. It shows both the occurrences of each circuit type, and the number of marriages involved in each circuit intersection.

***Classificatory Matrimonial Census – It is a specific kind of matrimonial census. It consists in searching not for individual circuit types, but for classes of circuit types, that share a given formal criterion.

***Closure Rate – In a kinship network, the closure rate indicates the probability that a marriage edge “close” a given kinship chain (thus, that this become a matrimonial circuit).

***Component – A component is a connected segment of a kinship network.

***Concordance Table - A concordance table is a file containing information about the renumbering of individuals. It is a simple text file with two columns where the Id numbers of the members of the second and the first corpus are listed. Individuals that appear in only one corpus do not have to be listed in this table, and numbers need not to be ordered.

***Connubial Circuit – It is a circuit as part of a matrimonial alliance network.

***Consanguine Density – It is the percentage of the pairs of individuals that are connected by a consanguine tie.

***Core - The core of a kinship network is the sum of its matrimonial bicomponents.

F

***Frame - The frame of a matrimonial circuit is the graph obtained by replacing each one of its consanguine chains with a simple arc. Consequently, only two types of lines compose it: the marriages and the consanguine ties. The frame of a matrimonial network is the union of all the circuit frames that compose the network.

G

***Generation - Generation and generational distance are not unique concepts. Except in kinship networks that consist of trees, there are usually several alternative ways to arrange an individual on a generational level inferior to its ascendants and superior to its descendants. For instance, if a man has married his sister's daughter, his children will be at the same time grandchildren and great-grandchildren of his father. One has to decide on the path along which generational distance shall be calculated.
The algorithm used by Puck is identical to that of Pajek. It consists in navigating through the network along kinship paths and assigning to parents, spouses and children of each individual the generational level of that individual, augmented by 1, 0 or -1 according to the nature of the kinship tie.
Note: The identity of the algorithm does not necessarily imply the network of the results. The result depends on the navigation path, which may be different for Pajek and Puck, since arcs are not necessarily stored in the same order.

K

***Kernel - The kernel of a kinship network is its largest matrimonial bicomponent.

***Kinship Network - "Kinship networks are characterized by the interplay of three fundamental principles: filiation, marriage, and gender. We ordinarily represent filiation by a set of arcs (descent arcs) that are directed from parents to children, and marriage by a set of undirected edges (marriage edges) between spouses (for alternative representations of kinship networks without edges, see below). Kinship networks thus are mixed graphs, containing both arcs and edges. Gender is usually taken into account by a partitioning of the vertex set (the gender partition), usually into two or three disjoint classes (male, female, and possibly unknown sex)" (Hamberger, Houseman, & White, 2012).

L

Linear Kinship Chain - A kinship chain is linear if it consists of uniformly oriented arcs.

M

***Matrimonial Alliance Network – It is a network of marriage ties between groups (classes) of individuals. In a matrimonial alliance network, arcs represent (one or more) marriage ties and point from the wife's to the husband's group ; nodes represent groups of individuals, who share a given property (place of residence, consanguinity...).

Matrimonial Bicomponent - A matrimonial bicomponent is a maximal subgraph in which every two vertices are part of a matrimonial circuit. Also, any two vertices in a matrimonial bicomponent can be linked to each other by two distinct kinship chains that do not pass through and do not meet in “structural children”. Matrimonial bicomponents are closely related (but not identical) to matrimonial components : both are line-biconnected (two distinct line-series link each vertex to every other), but matrimonial bicomponents are also vertex-biconnected (the two interconnecting line-series never run through the same vertex).

***Matrimonial Census (Circuit Census) – It is the search and exhaustive count of Matrimonial Circuits that comply with some quantitative (size) or qualitative (formula) criteria.

Matrimonial Circuit -  A matrimonial circuit is a kinship chain that both is closed by a marriage and do not involve childless and unmarried individuals (Hamberger, Houseman, Daillant, White, & Barry, 2004).
Pragmatically, matrimonial circuit types correspond to types of consanguine marriage (between consanguine kin, such as between a man and his mother’s brother’s daughter) and types of affine “relinkings” incorporating one, two or more intermediary marriage ties.
- Consanguine marriages, that incorporate a single marriage tie (and a single consanguine kinship chain), form matrimonial circuits of “width” 1, e.g. a man marries his mother’s brother’s daughter.
- Relinkings incorporating two marriage ties (and two consanguine kinship chains), form matrimonial circuits of “width” 2, e.g. a man and his sister marry a sister and her brother, or a man marries his mother’s brother’s wife’s bother’s daughter.
- Relinkings incorporating three marriage ties (and three consanguine kinship chains) form matrimonial circuits of “width” 3, e.g. a man marries his mother’s brother’s wife’s bother’s daughter’s husband’s sister.

Matrimonial circuits are indicators of sociological constraints of matrimonial choice (rules, preferences and avoidances, opportunities) and of the dynamics of self-organization of the network. They have to be studied as a whole. For the concept of the matrimonial circuit, see Hamberger, Houseman, & White (2012, p. 539‑540).

Matrimonial Component - A matrimonial component is the maximal connected part of a matrimonial network.

***Matrimonial Constellation – It is the largest component of a matrimonial network frame.

***CORR? Matrimonial Network - A matrimonial network is a subgraph induced by matrimonial circuits. Matrimonial networks are line-induced and not vertex-induced subgraphs. This means that every line of the subgraph is part of a circuit (it is not enough that its endpoints are in a circuit). The matrimonial network derived from a set of matrimonial circuits found in a kinship network is thus simply the network composed of these circuits. It consists, in other word, of the matrimonially “interesting” regions of the original kinship network.
The connected parts of the matrimonial network (the matrimonial components) represent continuous regions of densely interconnected circuits, which may be studied from various perspectives.
On the one hand, we may suppose that the frequent occurrence of particular matrimonial patterns is correlated with other properties of the network region concerned (for instance social class, geographical region or historical period); we may then apply several partitions to the network in order to evaluate the degree to which partition clusters correspond to matrimonial components.
On the other hand, we may interpret the density of circuits as an effect of self-reinforcing social mechanisms (behavior transmission, imitation or the presence of rules) or as a simple network effect (rings combining to compose other circuits) which we did not consider when defining the criteria for our initial circuit search.
The concept of a matrimonial network is also meaningful in and of itself, independent of any particular circuit set. Even without being able to precisely identify all matrimonial circuits (without limits of size) which may exist in a kinship network, it is possible to determine which part of the network is composed of matrimonial circuits. The result is the absolute matrimonial network, the subgraph induced by all lines in the network which are in some circuit whatsoever. This absolute matrimonial network is equivalent to the sum of all matrimonial bicomponents. It corresponds to what has been called the “core” in a P-graph context (Grange & Houseman, 2010; White & Houseman, 1996) .
Every matrimonial network constitutes a network without tails (every vertex must have a degree greater than one) and without structural children (every vertex must have an outdegree greater than zero). However, the reverse is not the case. There may be networks where all vertices fulfil these two-degree criteria, but which nevertheless are not matrimonial, as they contain lines which do not form part of any matrimonial circuit. Filial triads (father, mother and child) or marriage ties connecting disjoint matrimonial components are instances of this.

Matrimonial Ring - A matrimonial ring is a specific kind of matrimonial circuit, that do not include other circuits.

***Maximal Genealogical Depth – In a kinship network, it is the genealogical distance from the farthest ascendant.

***Mean Genealogical Depth – It is the mean genealogical distance from apical ascendants, it is calculated with the Cazes formula (Cazes, & Cazes, 1996).

N

***Nucleus (Absolute Matrimonial Network) – The nucleus of a kinship network is the network composed by all its matrimonial circuits.

O

Ore-graphs - Named after the scandinavian mathematician Oystein Ore (1970), developed by Vladimir Batagelj and Andrej Mrvar. In an Ore graph, vertices represent individuals, arcs filial ties and edges marriages. Vertex-labels represent gender; two different types of lines represent paternal and maternal ties.

P

Parental Triad - A parental triad is a graph formed by three vertices and arcs pointing from two of them (the “parents”) to the third (the “child”).

Path -  A path is an alternating sequence of vertices and lines (arcs or edges), where all vertices are distinct. Using this notion presuppose an ego-centered view of kinship ties.

P-graph - Developed by Douglas White and Paul Jorion (White, & Jorion, 1992), used by the homonymous computer program p-graph. In a P-graph couples or unmarried individuals are represented by vertices, married individuals by gender labeled lines running from the couple in which they are partner to the couple of which they are born.
P-graphs have the advantages of being directed acyclic and of incorporating fewer lines and vertices, allowing semi-cycles (that correspond to matrimonial circuits in Ore-graphs) to be more easily detected. Note, however, that an individual who marries several times will be represented by several lines. Lines therefore have to be name-labeled in order to distinguish identity from siblingship.

R

***Relation Census – It is a census of kinship relations that: whether comply with some given quantitative (size) or qualitative (formula) criteria; or correspond to the criteria established in a matrimonial census.

S

***Structural Children - In a kinship network, structural children are individuals who do not have neither spouses nor children.

T

Tip-graph - Named after the research group TIP (Traitement Informatique de la Parenté), used by the macros of the Tip4Pajek series (2007). In a Tip-graph, filial and marriage ties are represented by arcs. All information on the type of tie and on gender is contained in line values. There are five types of lines:

- a marriage arc pointing from female to male,
- a filial arc pointing from female (mother) to female (daughter),
- a filial arc pointing from female to male (son),
- a filial arc pointing from male (father) to female,
- a filial arc pointing from male to male.
Because a Tip-Graph does not involve vertex labeling, it is a highly economical representation of a kinship network. Its major disadvantage is that it is not directed acyclic. Many analyses therefore require its being re-transformed into a conventional Ore-graph. To export a dataset in tip-graph format (as a pajek project file) the option “tip” has to be chosen.

V
***Virtual Individuals - Virtual individuals are individuals for whom no information is available except for their kinship relations (and perhaps their gender, if they are parents or spouses of existing individuals). They only serve to represent the common parents of full siblings.

Détails

Mis à jour : 2 janvier 2016

PUCK

Program for the Use and Computation of Kinship data

© Research Group TIP (Kinship and Computing)
Centre National de Recherche Scientifique, Paris
Distributed under CeCILL licence version 2 (http://www.cecill.info/)
Created 2007 by Klaus Hamberger
Developed by Klaus Hamberger, Christian Momon, Edoardo Savoia, Telmo Menezes and Éric Mermet
Visualization powered by KinOath (developed by Peter Withers) and Geneaquilts (developed by Anastasia Bezerianos, Pierre Dragicevic and Jean-Daniel Fekete)
http://www.kintip.net/
JVM = 1.7.0_91 24.91-b01
XMX = 298M/353M/910M
OS = Linux 3.16.0-38-generic 64bits

SUPPORT
write to Puck users mailinglist : Cette adresse e-mail est protégée contre les robots spammeurs. Vous devez activer le JavaScript pour la visualiser.

Cette partie du site est actuellement en construction...

This section of the site is currently under construction...

What is PUCK ?
Puck is a computer program for analyzing genealogical and other kinship related data. It has been conceived in continuity with the gradual introduction of computers in kinship studies since the 1970s.
Puck is a product of the research group TIP and an outcome of the project « Informatical treatment of kinship phenomena » (« Traitement Informatique des phénomènes de parenté », 2006- 2009) financed by the French National Research Agency (ANR).
For a presentation of the software, and of its theoretical issues, we suggest english readers to see Hamberger, Houseman, & Grange, (2014) ; french readers can refer to Hamberger, Houseman, & Grange, (2009).

What PUCK does ?
« A paradox is at the heart of studies on kinship: the marriage choices which, as the foundation of kinship systems, should be the primary object were neglected, especially in their empirical dimension. The anthropology of kinship began as an analysis of terminology and developed into a science of rules and norms. Today, it appears as a multifaceted research field dealing with representations and institutions, political strategies and symbolic operations. Still, the actual practices that generate matrimonial networks continue to occupy a marginal position » (Hamberger, Houseman, & Grange, 2009, p. 107). Kinship cannot be treated as a collection of isolated elements, the use of a computer processing of matrimonial practices have proved fundamental.
Starting from this premise, the software allows not only to make very fine and precise analyses of the matrimonial structures and configurations of genealogical networks, but also to consider its quality and biases: the data processing of kinship relationships is thus an opportunity to consider the theoretical issues and methodological choices, as it respects the individual way of the collection and organization of field data or records, to assess the quality of the dataset studied.
Puck accompanies the researcher throughout his work, from data input to final analysis. It is compatible with most commonly used formats (Excel, Gedcom, Pajek, etc.), and allows to import or export files in all these formats. Puck has been written in Java 1.6 and is continuously updated.

PUCK main functions

• Matrimonial and Relational Census
  Puck is the first software allowing to run a full matrimonial and relational census on a genealogical dataset. It identifies the matrimonial circuits that can be found in a kinship network, classifies them and assists the researcher through the analysis of the network topology. By producing the circuit intersection network of a given dataset, it highlights the actual combinations existing between different marriages types and thus allows to focus on the most structurally significant areas of the network. Those operations can help in distinguishing whether the frequency of given matrimonial structures is due to specific social norms (preferences, avoidances...) or is a mechanical effect of the network density.
  

• Diagnostics of Kinship Datasets
  Genealogical corpora produced by researchers are not neutral objects. Genealogies constructed from informants or documentary sources are often incomplete and androcentric, which implies an asymmetry of relations in the network. It is very important to detect errors of data collection or data entry to correct them, to know biases and be able to relativize the raw results (Barry & Gasperoni, 2008; Hamberger & Gargiulo, 2013; Hamberger, Houseman, & White, 2012, p. 546‑547). Before analyzing a family dataset, the first step is therefore to establish a profile of the network, from basic tools (count of the population and demographic composition, gender distribution, genealogical depth, density) to more complex ones (family completeness, gender bias), not only on the overall network, but also on specific parts of it. By partitioning the dataset and focusing on specific parts of the network it is then possible to refine the analysis.
  
  
• Kinship Networks Partitioning
  PUCK allows to partition a kinship network by choosing appropriate criteria. These can be individual (e.g., gender, age, birth period, occupation...) as well as family (e.g., union status, children number...) properties. The partitioning operations can show themselves fundamental, in order to refine and supplement qualitative analysis (dataset diagnosis) and/or a matrimonial census.
  
  
• Kinship Networks Simulation 
  Puck can not only treat actual kinship networks, it also provide refined tools for their simulation. The technique, recently diffused in kinship studies, consist in producing randomly generated networks and using them as means of comparison. Such an analytical process can contribute to discern if some of the regularities found in matrimonial practices are the mere product of given statistical criterias or if they actually derive from social institutions.
  
  
• Management and Editing of Kinship Networks
  Hardware supports, such as fieldwork notebooks, hand-drawed genealogy graphs (etc.) are especially subject to time, weather and, more commonly, entropy. A 10 minutes rain can severely damage the fruit of a long period work. PUCK can be used as a tool to generate, manage and stock kinship networks on a software support. This can be crucial in order to secure the kinship data conservation and therefore their analysis.
  
  
• Navigation through Kinship Networks
  Kinship networks are complex collections of individuals, ties, events and properties. Even if such data have been gathered personally by the researcher, individual after individual, they can be surprisingly hard to remember and confusing to browse. Thereafter, supports such as fieldwork notebooks can carry biases : for instance, a household based enquiry can easily give unwanted prominence to the residential factor on other ones, like matrimonial alliances.
  Thus, Puck has been designed to provide researchers with a tool, not only for stocking data, but also for navigating through them as freely and neutrally as possible. Its components allow to move smoothly, as well as actually "jump", through kinship corpora.

 Download and System Requirements
The software can be downloaded here for free, where you will find the latest version (PUCK 2.0) and earlier ones (PUCK 1.0).

System requirements : written in Java, Puck works with most of the common operating systems : on PC (Linux, Solaris, Windows) as well as Mac. Before running the application, make sure that your computer is equipped with an adequate version of Java.

License
PUCK is a free software, distributed under CeCILL license (a french variant of GPL). Puck has been deposited at the APP (Agence pour la Protection des Programmes, Program Protection Agency) and is protected by French law.

Disclaimer
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS « AS IS » AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

About this guide
This guide has been conceived for users of every level. Its main objective is to offer accessible and clear answers to your questions about Puck 2, its functionalities and use. As the guide focuses on the program itself, we suggest exploring the bibliographic section in order to deepen the theoretical issues that surround the program conception.

In order to facilitate the access to the guide you can find, as follows, a short presentation of its different sections :

• The section called Functionalities contains a detailed and exhaustive description of PUCK functional components. It is the core of the guide. Its organisation follows the program components themselves and a specific menu is dedicated to it. In its "Start" page you can find a map of the program, indicating PUCK main components that are described in each page of the Functionalities section.
  
  
• The Alphabetic and Thematic indexes are meant to drive the reader through the guide contents. In particular, the Thematic Index is conceived to introduce in a more progressive way, new Puck users to the program.
  
  
• The Appendices section contains several additional chapters which contents do not directly concern Puck components themselves, but are useful introductions for its use. These are about :
  • How to Establish a Kinship Dataset
  • File Formats for Kinship Data
  • Kinship Relations Notation
  • Property Codes
  • Bibliographic References
    
    
• The Glossary offers a terminological compound to the guide reading. Most of its entries are an english translation of a previous glossary, that the TIP research group published in a monographic number of the french review Annales de démographie historique (2008/2 n.116, p.233-235).
  

Contact
For any issue with the program, please write to Puck users mailinglist : Cette adresse e-mail est protégée contre les robots spammeurs. Vous devez activer le JavaScript pour la visualiser.