This thesis consists of the four papers listed below and a survey of the research area. I Lina J. Andrén: Avoiding (m, m, m)-arrays of order n = 2k II Lina J. Andrén: Avoidability of random arrays III Lina J. Andr´en: Avoidability by Latin squares of arrays with even order IV Lina J. Andrén, Carl Johan Casselgren and Lars-Daniel Öhman: Avoiding arrays of odd order by Latin squares Papers I, III and IV are all concerned with a conjecture by Häggkvist saying that there is a constant c such that for any positive integer n, if m ≤ cn, then for every n × n array A of subsets of {1, . . . , n} such that no cell contains a set of size greater than m, and none of the elements 1, . . . , n belongs to more than m of the sets in any row or any column of A, there is a Latin square L on the symbols 1, . . . , n such that there is no cell in L that contains a symbol that belongs to the set in the corresponding cell of A. Such a Latin square is said to avoid A. In Paper I, the conjecture is proved in the special case of order n = 2k . Paper III improves on the techniques of Paper I, expanding the proof to cover all arrays of even order. Finally, in Paper IV, similar methods are used together with a recoloring theorem to prove the conjecture for all orders. Paper II considers another aspect of the problem by asking to what extent way a deterministic result concerning the existence of Latin squares that avoid certain arrays can be used when the sets in the array are assigned randomly. / Denna avhandling inehåller de fyra nedan uppräknade artiklarna, samt en översikt av forskningsområdet. I Lina J. Andrén: Avoiding (m, m, m)-arrays of order n = 2k II Lina J. Andrén: Avoidability of random arrays III Lina J. Andrén: Avoidability by Latin squares of arrays with even order IV Lina J. Andrén, Carl Johan Casselgren and Lars-Daniel Öhman: Avoiding arrays of odd order by Latin squares Artikel I, III och IV behandlar en förmodan av Häggkvist, som säger att det finns en konstant c sådan att för varje positivt heltal n gäller att om m ≤ cn så finns för varje n × n array A av delmängder till {1, . . . ,n} sådan att ingen cell i A i innehåller fler än m symboler, och ingen symbol förekommer i fler än m celler i någon av raderna eller kolumnerna, så finns en latinsk kvadrat L sådan att ingen cell i L innehåller en symbol som förekommer i motsvarande cell i A. En sådan latinsk kvadrat sägs undvika A. Artikel I innehåller ett bevis av förmodan i specialfallet n = 2k. Artikel III använder och utökar metoderna i Artikel I till ett bevis av förmodan för alla latinska kvadrater av jämn ordning. Förmodan visas slutligen för samtliga ordningar i Artikel IV, där bevismetoden liknar den som finns i i Artikel I och III tillsammans med en omfärgningssats. Artikel II behandlar en annan aspekt av problemet genom att undersöka vad ett deterministiskt resultat om existens av latinska kvadrater som undviker en viss typ av array säger om arrayer där mängderna tilldelas slumpmässigt.

Construction claims are considered by many project participants as one of the most disruptive and unpleasant events of a project (Ho & Liu, 2004). Researchers like Kumaraswamy (1996) argue that claims managers should focus not merely on the significant claims categories but also on the avoidable ones, to minimize the damaging effects on a given project. It becomes apparent that the causes underlying different claims categories have to be identified so that controllability and avoidability may be established. This research examines the issues by looking at the construction industry in the United Arab Emirates (UAE), the second largest single industry after oil and gas sector, where claims and disputes are a major problem in the country. 51 leading construction participants including clients, consultants and contractors alike in the UAE were sampled for the study. The analytical results of the survey were further examined, compared and validated via the analysis of the data collected from 45 construction projects. The main findings of the first phase of study lead to the establishment of an index system called Claim Focus Index (CFI) that further establishes the finding of the 16 most significant types of claims and disputes in the UAE. The study continues to examine the underlying causes of the most significant types of claims and disputes and has identified unique sets of the root causes specific to each significant type of claims and disputes. These shall establish the basis to formulate strategies to focus on avoidability and minimization of claims and disputes.

Dans ce document, nous étudions l’évitabilité de différentes formes de répétitions dans les mots. En particulier 3 des 6 chapitres sont dédiés aux répétitions abéliennes en lien notamment avec deux questions d’Erdős de 1957 et 1961. Nous commençons par montrer qu’il existe un algorithme décidant, sous certaines conditions, si un mot morphique évite des puissances abéliennes. Cet algorithme élargit la classe sur laquelle les précédents algorithmes pouvaient décider. Une généralisation de cet algorithme nous permet de montrer que les longs carrés abéliens sont évitables sur l’alphabet ternaire et que les carrés additifs sont évitables sur Z2 . Le premier résultat répond à une question ouverte de Mäkelä datant de 2003 alors que le deuxième rappelle la question ouverte de 1994 concernant l’évitabilité des carrés additifs sur Z.Une autre généralisation de notre algorithme permet d’étudier l’évitabilité des motifs au sens abélien. Nous montrons que les motifs binaires de longueur supérieure à 14 sont évitables sur l’alphabet binaire, améliorant la précédente borne de 118.Nous donnons des conditions suffisantes pour qu’un morphisme soit sans longues puissances nème k-abéliennes. Ce résultat nous permet de calculer, pour tout k ≥ 3, le nombre minimum de carrés k-abéliens qu’un mot binaire infini doit contenir en facteur. Il permet aussi de montrer que les longs carrés 2-abéliens sont évitables sur l’alphabet binaire et qu’il existe un mot ternaire qui ne contient qu’un seul carré 2-abélien en tant que facteur.Enfin, nous proposons une classification complète des formules binaires en fonction de la taille d’alphabet qu’il faut pour les éviter et du taux de croissance (exponentiel ou polynomial) du langage les évitant. / In this document, we study the avoidability of different kind of repetitions in words. We firstshow that under some conditions one can decide whether a morphic word avoids abelian n-thpowers. This algorithm can decide over a wider class of morphism than the previousalgorithms. We generalize this algorithm and use it to show that long abelian squares areavoidable over the ternary alphabet and that additive squares are avoidable over Z2 . The firstresult answers a weak version of a question formulated by Mäkelä in 2003 and the second oneis related to an open question from 1994 about the avoidability of additive squares over Z.Another generalization of this algorithm can be used to study avoidability of patterns in theabelian sense. In particular, we show that binary patterns of length more than 14 areavoidable over the binary alphabet in the abelian sense. This improves considerably theprevious bound of 118.We give sufficient conditions for a morphism to be long k-abelian n-th power-free. This resultallows us to compute for every k ≥ 3 the number of different k-abelian squares that a binaryword must contain. We prove that long 2-abelian squares are avoidable over the binaryalphabet and that over the ternary alphabet there exists a word that contains only one 2-abelian square.We also give a complete classification of binary formulas based on the size of the smallestalphabet over which they are avoidable and on the growth (exponential or polynomial) of theassociated language.

Les infections nosocomiales sont définies comme des infections contractées lors d’un séjour dans un établissement de soins. En matière de sécurité sanitaire, celles-ci sont déterminées par le principe de prévention et de précaution. En droit français, les infections nosocomiales ont longtemps été indemnisées en conséquences de solutions prétoriennes. Ce régime indemnitaire repose aujourd'hui sur le principe de la responsabilité sans faute qui a été consacré par la loi du 4 mars 2002. Toutefois la difficile prise en charge de cette indemnisation par les assureurs conduisit à l’adoption d’un nouveau régime découlant de la loi du 30 décembre 2002. Ces lois successives ont rendu complexe et ambigu le mécanisme de la réparation des dommages causés par les infections nosocomiales. Le constat résulte du fait qu’il manque une définition juridique de l’infection nosocomiale, outil essentiel à la pérennité du système indemnitaire mis en place pour la réparation des dommages causés par ces infections. / Nosocomial infections are infections acquired during a stay in a health care institution. From the sanitary safety point of view, they are defined by the prevention principle and the precautionary principle. According to the French law, the nosocomial infections have long been compensated by a case law. Nowadays, this compensation system is based on the principle of liability without fault laid down by the law of the 4th of March 2002. However, the reluctance of the insurers to assume this compensation led to the adoption of a new system resulting from the law of the 30th of December 2002. Because of these successive laws, the compensation mechanism for the damages due to nosocomial infections has been made more complex and ambiguous. This observation stems from the lack of a legal definition of the nosocomial infection, an essential tool for ensuring the sustainability of the compensation system for the damages caused by these infections

<p>Кривица постоји ако је учинилац у време када је учинио кривично дело био урачунљив и поступао са умишљајем (нехат кад је законом одређено), а био је свестан или је био дужан и могао бити свестан да је његово дело забрањено. Другим речима, учинилац који није знао да је оно што ради забрањено и који није могао (нити био дужан) да то зна, није крив. Уколико учинилац није знао да је његово дело забрањено, а (био је дужан и) могао је да зна, казна му се може ублажити. Правна заблуда је институт кривичног права уско повезан са начелом кривице, а ипак, као да није успео да се профилише на начин који му припада. Уколико лице нема свест о забрањености дела, налази се у правној заблуди. Она је супротан пол свести о забрањености дела, а свест о забрањености дела је, према чл. 22 КЗ Србије, једна од три компоненте кривице, уз урачунљивост и умишљај (и нехат када је то законом одређено). Три битна обележја савременог кривичног права су хиперкриминализација, учесталост измена и допуна и бланкетизација.У КЗ из 1951. године било је прописано 216 кривичних дела, у КЗ из 1977. године 227, а у КЗ из 2006. године 326 кривичних дела. Од 2009. године до 2015. године донето је 15 закона у којима је прописано 33 кривична дела. КЗ из 1977. године се у 26 година мењао 25 пута (1977. до 2003.), а КЗ из 2006. године се за 8 година (2006. до 2014.) мењао 8 пута. У кривичном законодавству Републике Србије прописано је укупно 385 кривичних дела (основно 326 кривичних дела, а споредно законодавство 59). Од укупног броја кривичних дела, 261 има у потпуности или делимично бланкетни карактер, односно 67,8 %. Бланкетна кривична дела садрже један или више појмова чије значење је одређено у неком другом пропису. Увидом у 326 кривичних дела прописаних у КЗ, утврдили смо да бланкетна кривична дела упућују на 155 прописа. Ова бројка свакако није коначна, напротив. Нису урачунати сви они &bdquo;невидљиви&ldquo; бланкети попут: &bdquo;општепризнатих техничких правила&ldquo;; &bdquo;прописа и мера&ldquo;; &bdquo;техничких правила о мерама заштите&ldquo; и др. У раду се прво указује на одређене појаве које &bdquo;подстичу озбиљнији приступ&ldquo; разматрању института правне заблуде. Ради се о хиперкриминализацији, те неједнакој опасности понашања која су прописана као кривична дела. Након тога следе излагања о појму и развоју кривице (у кривичном праву), кроз кратку дискусију о слободи воље; а потом се приказују становишта психолошких и нормативних теорија кривице. На крају тог наслова објашњава се шта су основи искључења кривице, шта основи извињавања и у чему се разликују. Истраживање о правној заблуди почиње темом свести о забрањености дела која обухвата: садржај, дељивост, условљеност и место свести о противправности; могућности вештачења ове свести те појам и историјски развој правне заблуде. Након тога излаже се део о врстама правне заблуде који обухвата и обрнуту правну заблуду и њене појавне облике. Следећи део представља тему која обрађује &bdquo;прерушавање правне у стварну заблуду&ldquo;. Ради се о низу кривичних дела у чијим законским описима се појављује обележје противправности. Расправља се подела противправности из законског описа дела као општег и посебног обележја дела. Затим, анализа отклоњивости правне заблуде обухвата: критеријуме и средства отклоњивости заблуде те низ конкретних случајева када је правна заблуда (не)отклоњива на примерима из судске праксе Немачке и Аустрије који су подељени у четири целине и коментаре 45 утврђених схватања о отклоњивости. Следе излагања о утицају бланкетних кривичних дела на правну заблуду и односу стварне и правне заблуде код ових дела, као и дејству правне заблуде данас и у прошлости. Учињен је осврт и на упоредноправна решења. На крају се обрађују питања стварне заблуде, нарочито у ширем смислу, уз теорије које настоје да објасне дејство ове заблуде, као и заблуда о извињавјућим основима. Рад завршава напоменама о (могућем) поступању судова у примени овог института као и списком бланкетних прописа, те закључним разматрањима.</p> / <p>Krivica postoji ako je učinilac u vreme kada je učinio krivično delo bio uračunljiv i postupao sa umišljajem (nehat kad je zakonom određeno), a bio je svestan ili je bio dužan i mogao biti svestan da je njegovo delo zabranjeno. Drugim rečima, učinilac koji nije znao da je ono što radi zabranjeno i koji nije mogao (niti bio dužan) da to zna, nije kriv. Ukoliko učinilac nije znao da je njegovo delo zabranjeno, a (bio je dužan i) mogao je da zna, kazna mu se može ublažiti. Pravna zabluda je institut krivičnog prava usko povezan sa načelom krivice, a ipak, kao da nije uspeo da se profiliše na način koji mu pripada. Ukoliko lice nema svest o zabranjenosti dela, nalazi se u pravnoj zabludi. Ona je suprotan pol svesti o zabranjenosti dela, a svest o zabranjenosti dela je, prema čl. 22 KZ Srbije, jedna od tri komponente krivice, uz uračunljivost i umišljaj (i nehat kada je to zakonom određeno). Tri bitna obeležja savremenog krivičnog prava su hiperkriminalizacija, učestalost izmena i dopuna i blanketizacija.U KZ iz 1951. godine bilo je propisano 216 krivičnih dela, u KZ iz 1977. godine 227, a u KZ iz 2006. godine 326 krivičnih dela. Od 2009. godine do 2015. godine doneto je 15 zakona u kojima je propisano 33 krivična dela. KZ iz 1977. godine se u 26 godina menjao 25 puta (1977. do 2003.), a KZ iz 2006. godine se za 8 godina (2006. do 2014.) menjao 8 puta. U krivičnom zakonodavstvu Republike Srbije propisano je ukupno 385 krivičnih dela (osnovno 326 krivičnih dela, a sporedno zakonodavstvo 59). Od ukupnog broja krivičnih dela, 261 ima u potpunosti ili delimično blanketni karakter, odnosno 67,8 %. Blanketna krivična dela sadrže jedan ili više pojmova čije značenje je određeno u nekom drugom propisu. Uvidom u 326 krivičnih dela propisanih u KZ, utvrdili smo da blanketna krivična dela upućuju na 155 propisa. Ova brojka svakako nije konačna, naprotiv. Nisu uračunati svi oni &bdquo;nevidljivi&ldquo; blanketi poput: &bdquo;opštepriznatih tehničkih pravila&ldquo;; &bdquo;propisa i mera&ldquo;; &bdquo;tehničkih pravila o merama zaštite&ldquo; i dr. U radu se prvo ukazuje na određene pojave koje &bdquo;podstiču ozbiljniji pristup&ldquo; razmatranju instituta pravne zablude. Radi se o hiperkriminalizaciji, te nejednakoj opasnosti ponašanja koja su propisana kao krivična dela. Nakon toga slede izlaganja o pojmu i razvoju krivice (u krivičnom pravu), kroz kratku diskusiju o slobodi volje; a potom se prikazuju stanovišta psiholoških i normativnih teorija krivice. Na kraju tog naslova objašnjava se šta su osnovi isključenja krivice, šta osnovi izvinjavanja i u čemu se razlikuju. Istraživanje o pravnoj zabludi počinje temom svesti o zabranjenosti dela koja obuhvata: sadržaj, deljivost, uslovljenost i mesto svesti o protivpravnosti; mogućnosti veštačenja ove svesti te pojam i istorijski razvoj pravne zablude. Nakon toga izlaže se deo o vrstama pravne zablude koji obuhvata i obrnutu pravnu zabludu i njene pojavne oblike. Sledeći deo predstavlja temu koja obrađuje &bdquo;prerušavanje pravne u stvarnu zabludu&ldquo;. Radi se o nizu krivičnih dela u čijim zakonskim opisima se pojavljuje obeležje protivpravnosti. Raspravlja se podela protivpravnosti iz zakonskog opisa dela kao opšteg i posebnog obeležja dela. Zatim, analiza otklonjivosti pravne zablude obuhvata: kriterijume i sredstva otklonjivosti zablude te niz konkretnih slučajeva kada je pravna zabluda (ne)otklonjiva na primerima iz sudske prakse Nemačke i Austrije koji su podeljeni u četiri celine i komentare 45 utvrđenih shvatanja o otklonjivosti. Slede izlaganja o uticaju blanketnih krivičnih dela na pravnu zabludu i odnosu stvarne i pravne zablude kod ovih dela, kao i dejstvu pravne zablude danas i u prošlosti. Učinjen je osvrt i na uporednopravna rešenja. Na kraju se obrađuju pitanja stvarne zablude, naročito u širem smislu, uz teorije koje nastoje da objasne dejstvo ove zablude, kao i zabluda o izvinjavjućim osnovima. Rad završava napomenama o (mogućem) postupanju sudova u primeni ovog instituta kao i spiskom blanketnih propisa, te zaključnim razmatranjima.</p> / <p>The Guilt exists if the perpetrator at the time he committed the offense was countable and acted with intent (negligence when it was determined by law), and was aware of or was obliged and could be aware that his act was prohibited. In other words, an offender who did not know that what he was doing was prohibited and who could not (nor was obliged to) know it, is not guilty. If the perpetrator did not know that his action was prohibited, and (he was obliged) he could have known, his punishment could be mitigated. Mistake of law is a criminal law institute closely linked to the principle of guilt, and yet, as if it did not manage to profil the way it belongs. If a person has no awareness of the prohibition of act, he is in a mistake of law. It is the opposite pole of the conscience of the prohibition of the act, and the conscience of the prohibition of the action is, according to Art. 22 of the CC of Serbia, one of the three components of guilt, with accountability and intent (and negligence when determined by law). Three important features of modern criminal law are overcriminalization, frequency of amendments and blanketization. In the CC from 1951, 216 criminal offences were prescribed, in the CC from 1977, 227, and in the CC from 2006, 326 criminal offences. From 2009 to 2015, 15 laws were passed in which 33 criminal offences were prescribed. The CC from 1977 changed 25 times in 26 years (1977 to 2003), and the 2006 CC changed 8 times in 8 years (2006 to 2014). The criminal legislation of the Republic of Serbia consists of&nbsp; 385 criminal offences (326 criminal offences in basic, and 59 in secondary legislation). Out of the total number of crimes, 261 has a completely or partially blanket character, that is, 67.8%. Blanket criminal offences contain one or more terms the meaning of which is defined in another regulation. By inspecting the 326 criminal offences prescribed in the Criminal Code, we established that blanket offences refer to 155 regulations. This figure is certainly not the final, on the contrary. All those &quot;invisible&quot; blankets are not counted as: &quot;generally accepted technical rules&quot;; &quot;Regulations and measures&quot;; &quot;Technical rules on protection measures&quot; and others. The text first points to certain phenomena which &quot;encourage a more serious approach&quot; to the consideration of the institute of mistake of law. This is an overcriminalization, and an unequal risk of behaviors that are prescribed as criminal offences. After that, there follows the statements about the concept and the development of guilt (in criminal law), through a brief discussion of freedom of will; and then the views of psychological and normative theories of guilt are presented. At the end of this title, the explanation is made regarding the differnece between basis of the excuses of guilt and apology. A study of mistake of law begins with the subject of conscience of the prohibition of action that includes: content, divisiveness, conditionality and place of conscience of the prohibition; possibilities for expert examination of this conscience, and the concept and historical development of mistake of law. Thereafter, a section on the types of mistake of law, which includes both reversed mistake of law and its manifestations, is presented. The next section presents a topic that deals with &quot;disguising the mistake of law in a mistake of fact&quot;. It is about series of criminal offences in whose legal descriptions a mark of unlawfulness appears. The division of illegality from the legal description of the work as a general and special feature of the action is discussed. Furthermore, the analysis of the avoidability of the mistake of law includes: the criteria and means of avoidability of the mistake of law and a number of cases where the mistake of law is (not) avoidable on examples from the case law of Germany and Austria divided into four parts and 45 comments of the established perceptions of the avoidability. Following are the statements about the impact of blanket crimes on mistake of law and the relation of mistake of law and fact to these acts, as well as to the legal effect of mistake of law today and in the past. A review was also made of comparative solutions. In the end, questions of mistake of fact, especially in the wider sense, are addressed, along with theories that attempt to explain the effect of this mistake, as well as the mistake of apologizing basis. The text concludes with remarks on the (possible) conduct of courts in the application of this institute, as well as the list of blank regulations, and concluding observations.</p>

Nous étudions dans cette thèse différents problèmes d'évitabilité des répétitions dans les mots infinis. Soulevée par Thue et motivée par ses travaux sur les mots sans carrés, la problématique s'est développée au cours du XXe siècle, et est aujourd'hui devenue un des grands domaines de recherche en combinatoire des mots. En 1972, Dejean proposa une importante conjecture, dont la validation étape par étape s'est terminée récemment (2009). La conjecture concerne le seuil des répétitions d'un alphabet, i.e., la borne inférieure des exposants évitables sur cet alphabet. La notion de seuil, comme frontière entre évitabilité et non-évitabilité d'un ensemble donné de mots, est le fil directeur de nos travaux. Nous nous intéressons d'abord à une généralisation du seuil des répétitions (nous donnons des encadrements de sa valeur). Cette notion permet d'ajouter, pour décrire l'ensemble des répétitions à éviter, au paramètre de l'exposant, celui de la longueur des répétitions. Puis, nous étudions des problèmes d'existence de mots dans lesquels, simultanément, certaines répétitions sont interdites et d'autres sont forcées. Nous répondons, pour l'alphabet ternaire, à la question : quels réels sont l'exposant critique d'un mot infini sur un alphabet fixé? Nous introduisons ensuite une notion de haute répétitivité, et établissons une description partielle des couples d'exposants paramètrant une double contrainte de haute répétitivité et d'évitabilité. Pour finir, nous utilisons des résultats et techniques issus de ces problématiques pour résoudre une question de coloration de graphes : nous introduisons un seuil des répétitions, calqué sur celui connu pour les mots, et donnons sa valeur pour deux classes de graphes, les arbres et les graphes de subdivisions. / In this thesis we study various problems on repetition avoidance in infinite words. Raised by Thue and motivated by his work on squarefree words, the topic developed during the 20th century, and has nowadays become a principal area of research in combinatorics on words. In 1972, Dejean proposed an important conjecture whose verification in steps was completed recently (2009). The conjecture concerns the repetition threshold for an alphabet, i.e., the infimum of the avoidable exponents for that alphabet. The notion of threshold as a borderline between avoidability and unavoidability for a given set of words is the guiding line of our work. First, we focus on a generalization of the repetition threshold. This concept allows us to include, in addition to the exponent, the length of the repetitions as a parameter in the description of the set of repetitions to avoid. We obtain various bounds in that respect. We then study existence problems for words in which simultaneously some repetitions are forbidden, and others are forced. For the ternary alphabet, we answer the question: what real numbers are the critical exponent of some infinite word over a given alphabet? Also, we introduce a notion of highly repetitive words and give a partial description of the pairs of exponents which parameterize the existence of words both highly repetitive and repetition-free. Finally, we use results and techniques stemming from those problems to solve a question on graph colouring: we introduce a repetition threshold adapted from the thresholds we know for words, and give its value for two classes of graphs, namely, trees and subdivision graphs.

Bibliography: pages 192-195 / The avoidability, or unavoidability of patterns in words over finite alphabets has been studied extensively. The word α over a finite set A is said to be unavoidable for an infinite set B+ of nonempty words over a finite set B if, for all but finitely many elements w of B+, there exists a semigroup morphism φ ∶ A+ → B+ such that φ(α) is a factor of w. In this treatise, we start by presenting a historical background of results that are related to unavoidability. We present and discuss the most important theorems surrounding unavoidability in detail. We present various complexity-related properties of unavoidable words. For words that are unavoidable, we provide a constructive upper bound to the lengths of words that avoid them. In particular, for a pattern α of length n over an alphabet of size r, we give a concrete function N(n, r) such that no word of length N(n, r) over the alphabet of size r avoids α. A natural subsequent question is how many unavoidable words there are. We show that the fraction of words that are unavoidable drops exponentially fast in the length of the word. This allows us to calculate an upper bound on the number of unavoidable patterns for any given finite alphabet. Subsequently, we investigate computational aspects of unavoidable words. In particular, we exhibit concrete algorithms for determining whether a word is unavoidable. We also prove results on the computational complexity of the problem of determining whether a given word is unavoidable. Specifically, the NP-completeness of the aforementioned problem is established. / Decision Sciences / D. Phil. (Operations Research)

The characterization of the Markan disciples has been and continues to be the object of much scholarly reflection and speculation. For many, the Markan author's presentation of Jesus' disciples holds a key, if not the key, to unlocking the purpose and function of the gospel as a whole. Commentators differ as to whether the Markan disciples ultimately serve a pedagogical or polemical function, yet they are generally agreed that the disciples in Mark come off rather badly, especially when compared to their literary counterparts in Matthew, Luke, and John. This narrative-critical study considers the characterization of the Markan disciples within the Sea Crossing movement (Mark 4:1-8:30). While commentators have, on the whole, interpreted the disciples' negative characterization in this movement in terms of lack of faith and/or incomprehension, neither of these, nor a combination of the two, fully accounts for the severity of language leveled against the disciples by the narrator (6:52) and Jesus (8:17-18). Taking as its starting point an argument by Jeffrey B. Gibson (1986) that the harshness of Jesus' rebuke in Mark 8:14-21 is occasioned not by the disciples' lack of faith or incomprehension but by their active resistance to his Gentile mission, this investigation uncovers additional examples of the disciples' resistance to Gentile mission, offering a better account of their negative portrayal within the Sea Crossing movement and helping explain many of their other failures. In short, this study argues that in Mark 4:1-8:26, the disciples are characterized as resistant to Jesus' Gentile mission and to their participation in that mission, the chief consequence being that they are rendered incapable of recognizing Jesus' vocational identity as Israel's Messiah (Thesis A). This leads to a secondary thesis, namely, that in Mark 8:27-30, Peter's recognition of Jesus' messianic identity indicates that the disciples have finally come to accept Jesus' Gentile mission and their participation in it (Thesis B). Chapter One: Introduction: offers a selective review of scholarly treatments of the Markan disciples, which shows that few scholars attribute resistance, let alone purposeful resistance, to the disciples. Chapter Two: The Rhetoric of Repetition: introduces the methodological tools, concepts, and perspectives employed in the study. It includes a section on narrative criticism, which focuses upon the story-as-discoursed and the implied author and reader, and a section on Construction Grammar, a branch of cognitive linguistics founded by Charles Fillmore and further developed by Paul Danove, which focuses upon semantic and narrative frames and case frame analysis. Chapter Three: The Sea Crossing Movement, Mark 4:1-8:30: addresses the question of Markan structure and argues that Mark 4:1-8:30 comprises a single, unified, narrative movement, whose action and plot is oriented to the Sea of Galilee and whose most distinctive feature is the network of sea crossings that transport Jesus and his disciples back and forth between Jewish and Gentile geopolitical spaces. Following William Freedman, Chapter Four: The Literary Motif: introduces two criteria (frequency and avoidability) for determining objectively what constitutes a literary motif and provides the methodological basis and starting point for the analyses performed in chapters five and six. Chapter Five: The Sea Crossing Motif: establishes and then carries out a lengthy narrative analysis of the Sea Crossing motif, which is oriented around Mark's use of ‎θάλασσα (thalassa) and πλοῖον (ploion), and Chapter Six: The Loaves Motif: does the same for The Loaves motif, oriented around Mark's use of ἄρτος (artos). Finally, Chapter Seven: The Narrative Logic of the Disciples (In)comprehension: draws together all narrative, linguistic, and exegetical insights of the previous chapters and offers a single coherent reading of the Sea Crossing movement that establishes Theses A and B.