Scholarly article on topic '‘Thrilled with Chilly Horror’: A Formulaic Pattern in Gothic Fiction'

‘Thrilled with Chilly Horror’: A Formulaic Pattern in Gothic Fiction Academic research paper on "Languages and literature"

Share paper
Academic journal
Studia Anglica Posnaniensia
OECD Field of science

Academic research paper on topic "‘Thrilled with Chilly Horror’: A Formulaic Pattern in Gothic Fiction"

Studia Anglica Posnaniensia 49/2, 2014 doi: 10.24 78/stap-2014-0010



Manuel Aguirre

Independent Scholar


This article is part of a body of research into the conventions which govern the composition of Gothic texts. Gothic fiction resorts to formulas or formula-like constructions, but whereas in writers such as Ann Radcliffe this practice is apt to be masked by stylistic devices, it enjoys a more naked display in the - in our modern eyes - less 'canonical' Gothics, and it is in these that we may profitably begin an analysis. The novel selected was Peter Teuthold's The Necromancer (1794) - a very free translation of K. F. Kahlert's Der Geisterbanner (1792) and one of the seven Gothic novels mentioned in Jane Austen's Northanger Abbey.

There is currently no literature on the topic of formulaic language in Gothic prose fiction. The article resorts to a modified understanding of the term 'collocation' as used in lexicography and corpus linguistics to identify the significant co-occurrence of two or more words in proximity. It also draws on insights from the Theory of Oral-Formulaic Composition, in particular as concerns the use of the term 'formula' in traditional epic poetry, though again some modifications are required by the nature of Teuthold's text. The article differentiates between formula as a set of words which appear in invariant or near-invariant collocation more than once, and a formulaic pattern, a rather more complex, open system of collocations involving lexical and other fields. The article isolates a formulaic pattern—that gravitating around the node-word 'horror', a key word for the entire Gothic genre -, defines its component elements and structure within the book, and analyses its thematic importance. Key to this analysis are the concepts of overpatterning, ritualization, equivalence and visibility.

Keywords: collocation, equivalence, field, formula, formulaic pattern, Gothic fiction, horror, overpatterning, ritualization, visibility

1. Introduction

This article is part of a body of research into the conventions which govern the composition of Gothic texts.1 One line of interest concerns the observed fact that Gothic fiction resorts to formulas or formula-like constructions; whereas in the work of writers such as Ann Radcliffe this practice is apt to be masked by stylistic devices, it enjoys a more naked display in those Gothics which our own times consider less 'canonical', and it is in these that we may profitably begin analysis. The novel selected was Peter Teuthold's The Necromancer (1794).2 Though it presents itself as an English rendering of K. F. Kahlert's Der Geisterbanner (1792), the translator's rhetoric differs considerably from that of his German original, and it is best to proceed on the assumption that we are dealing with an English book.3

In the historical sense in which I will use the term, 'Gothic' designates a circa-60-year period which 'began' with the publication of Horace Walpole's The castle of Otranto in 1764 and was superseded by other forms of horror fiction from around the 1820s onwards.4 The rise of the genre is a complex issue but it can usefully be attributed to, among other things, a reaction against facile Enlightenment positions; in many ways Gothic fiction may be viewed as the Romantic fiction which (with the occasional exceptions of Scott or Mary Shelley) is absent from so many anthologies of and critical works on Romantic literature. The genre (which, besides novels, includes drama, poetry and short fiction) capitalizes on strategies associated with the Graveyard School of Poetry, the sentimental novel, and generally the valorisation of the non-rational (feeling, the passions, the Burkean Sublime),5 but it also relates to a type of realism which, shunned by earlier fictions, dwelt on defeat or powerlessness in the face of forces greater than the enlightened will of the individual or of society. It is my contention that for a firm grasp of these issues a frank examination of the language and structure of Gothic is essential.

There is, so far as I know, no literature on the topic of Gothic formulaic language. For that matter, there is precious little critical study of the formal aspects of the Gothic genre, possibly owing to the misconception that, in terms of artis-

1 For the overall initiative see The Northanger Library Project (www.northangerlibrary .com); see also Aguirre 2013a, 2013b, and 2014.

All references will be to this edition. This is one of the seven Gothic titles mentioned by

Jane Austen in Northanger Abbey.

For some comparison between the two books see Hall 2000 and Murnane 2010. Justification for this choice is provided in Aguirre 1990, esp. p. 209.

For these and other sources see Punter 1980, Botting 1996; on the folk sources of Gothic see also Aguirre 2013a.

tic quality, the genre has little that will endear itself to us.6 For two-and-a-half centuries criticism has adhered to content-based definitions - motif, episode, plot, theme, setting, atmosphere, psychology, ideology - while none that I am aware of build on formal criteria - lexicon, formulaic language, syntax, style, narrative structure, and so on. The approach provided here is weary of a priori definitions of Gothic which consistently ignore the very forms of the genre; and the rationale of this article hinges on a formal approach.

The major tools for the present research were the theory of Oral-Formulaic Composition and insights from corpus linguistics and lexicography.7 It was noted long ago that formulaic constructions in traditional epic poetry rely on combinatorial strategies built on 'substitution systems',8 and research has shown that the work of Parry, Lord and others on formulaic diction has a genuine applicability to the Gothic genre, albeit a limited one inasmuch as their first, and defining, criterion is a metrical one; thus, for Parry (1930: 80) the formula is 'a group of words which is regularly employed under the same metrical conditions to express a given essential idea'. Though some phonological features of the epic formula (notably alliteration and stress) will disclose an unexpected relevance to the present research (see below), the concept of formulaic composition seemed intuitively appropriate but difficult to apply to a piece of 18th-century prose fiction.

On the other hand, the tendency in The Necromancer towards certain complex lexical groupings required some principle of correlation beyond (or before) syntactic structure for which the concept of collocation (see Firth 1957) initially offered sufficient precision. 'Collocation' identifies the frequent co-occurrence of two or more words in contiguity or proximity (in practice, most researchers handle word-pairs); the item selected as the focus for the analysis is called the node, the words the node tends to be accompanied by its collocates. It is often but by no means always the case that the words may be knit together by some syntactic arrangement (Sinclair 1991; Hoey 1991: 154). Soon, however, the limitations of this concept became clear. For one thing, lexicographers investigating formulaic language seem largely unaware of the Parry-Lord theory and over eighty years of critical research; for another, focus on the spoken language risks compromising the specificity of the written, literary text; furthermore, the tendency is to study hardened or lexicalized strings as the model for other types of string construction (e.g., Wray 2005), whereas the object of my investigation is a productive system for generating text.9

For an examination of critical positions in this regard see the first section in Aguirre 2013a.

One other important source of insight was the folklorists' studies of the fairytale formula. For details and references see Chapter 4 in Aguirre 2007.

Pioneering work in this field was conducted by Milman Parry (1930); see also Lord 1960, Fry 1967, and Foley 1988.

Collocation is currently viewed as the lowest level on a scale of co-occurrence relations in

For the purposes of this enquiry into Gothic procedures, the formula, understood as consisting of a set of words which appear in near-invariant collocation more than once, is to be distinguished from the formulaic pattern (the key term in this study). This latter is best defined as a polylexemic phenomenon (rather than as involving pairs of words) that binds lexical fields (rather than lexical items) around a node-field (rather than a node-word).10 No true constants are to be found in the formulaic pattern, which thus appears to be an open system (unlike the formula properly so called). Over and above syntactic considerations, textual units co-occur in it with other textual units through proximity and combination. Provisionally we can define the formulaic pattern in Teuthold's novel as an open set of lexical fields regularly employed under certain conditions to convey an essential idea. In order to tighten up this definition the article will isolate one single formulaic pattern, define its component elements and structure within the novel, and analyse its thematic importance.

2. The corpus

In The Necromancer, the word 'horror' appears fifty-seven times in the following syntactic positions: Subject (x4); Direct Object (x8); in the prepositional phrases 'with horror' (x23) and 'of horror' (x18); and in other prepositional constructions (x4). This article will concentrate on the largest of these sets, the phrase 'with horror'. As the novel has no chapters, examples are identified by volume and page (e.g., 1.27, 2.186). The language, spelling and punctuation of the original have been scrupulously respected. There are no fewer than ten narrators, and for reasons that will become clear later the narrator's name is added after each instance.

1.24 how was I chilled with horror when I beheld the features of my deceased mother! (Hellfried)

1.41 I staggered back thrilled with chilly horror [on hearing his mother's name ut-

tered by the Necromancer] (Hellfried)

1.51 when I suddenly recollected the mysterious words of our landlord, I cannot

but confess that I was seized with horror (Herrman)

language; the three other levels are colligation, semantic preference, and semantic prosody (see Hoey 2000, Louw 1993, Sinclair 2004). Whereas some of these notions have much relevance to the present investigation, the complexity of the issue (for one thing, Foley (1990: 238) makes collocation the first link in a very different chain - collocation, cluster, theme, and special rhetorical structures) makes it advisable to tackle it in a separate article. These preliminary results are presented in a companion article, "'The tranquillity of the mansion': Fields and formulaic diction in a Gothic novel" (Aguirre 2015).

1.65 The landlord ... was struck with horror and amazement when we related

where we had been (Herrman) 1.78 Our valiant crew was still fixed to the ground, seized with horror and aston-

ishment [on hearing mysterious sounds] (Herman) 1.89 it flew suddenly open, and we beheld with horror a black vault (Herman) 1.92 The phantom ... trembled violently as it rose, and emitted an hollow groan,

looking around with chilly horror (Herman) 1.103 The apprehension of being famished in that infernal abode, thrilled my soul

with horror, and black despair (Herrman) 1.134 a shiny figure . appeared . I shuddered, seized with horror, and was unable to utter a single word (the old widow, as reported by the Austrian officer, as reported by Lieutenant B -) 1.137 I tremble, seized with chilly horror, when my frantic mind anticipates that awful moment [of divine retribution] (the old widow, as reported by the Austrian officer, as reported by Lieutenant B -) 1.153 The vision stared at me ., it made my blood congeal, thrilling my soul with

deadly horror (the Austrian officer, as reported by Lieutenant B -) 1.155 [after the phantom's departure,] my comrades were chilled with horror (the

Austrian officer, as reported by Lieutenant B - ) 1.166 I was struck with surprise when I beheld Baron T - before me, and I cannot but confess that I was seized with horror, when I saw him (the Austrian officer, as reported by Lieutenant B -) 2.11 we beheld with chilly horror an human figure hovering on the wall (Lieuten-

ant B -)

2.11 After we had gazed at the phantom some time, with secret horror (Lieutenant

2.43 I made another discovery which thrilled my soul with horror: Perceiving

traces of blood on the pillow, I was seized with a sudden terror (Lieutenant B -)

2.106 His looks grew wild, ... and he heaved a deep groan; whilst I was standing

before him like a statue, thrilled with horror and pity (Lieutenant B -) 2.145 I shall not be able to give you a full narrative of deeds which fill my soul with

horror (Volkert, as reported by Lieutenant B -) 2.153 The phantom ... looked at the watch ..., gave him a menacing look, and left

him thrilled with horror and amazement (Volkert, as reported by Lieutenant B -)

2.170 The dismal idea of the awful scene [of Volkert's imminent execution] haunted me where ever I went .; my soul was pierced with horror when the first ray of the rising sun hailed me on my couch (Lieutenant B -) 2.172 I looked back and beheld with horror a thick column of smoke (Lieutenant B -)

2.182 My fellow prisoners sung obscene songs to me, which I could not hear without

disgust and horror (Wolf, as reported by Lieutenant B -) 2.187 I beheld with horror that she was infected with an ignominious illness (Wolf, as reported by Lieutenant B -)

3. Lexical fields

Aside from a few cases of exact repetition ('beheld with horror', 'with chilly horror'), the items offer various types and degrees of resemblance; a 'family air' rules in that we seem to come across the 'same' words and constructions all the time, but no classification procedure suggests itself in any obvious way. To uncover some principle of order we will begin by examining the lexicon used. The word 'horror' in this corpus is part of a lexical field (LF1) - an open paradigm of emotion terms such as 'terror' or 'astonishment'. The adjectives bearing on these ('chilly', 'deadly', etc.) shape their own lexical field (LF2) central to which stand notions of the concealed, the unexpected and the threatening. Another field (LF3) consists of the set of possible individuals undergoing the experience of horror. Next we have a field containing terms of perception (LF4); these include verbs in the active voice ('beheld', 'gazed', etc.), while in many additional cases the context leaves us in no doubt as to the perceptual source of the experience: horror in this corpus emerges from sight (x14), hearing (x3), expectation (x3) (which at least in one case (2.170) has to do with the anticipation of a 'scene'), recollection (x2) and narrative heard (1.65). Next we have (LF5a) transitive verbs of physical impact such as 'thrilled' (x6) or 'seized' (x5), mostly in the passive voice. Another group (LF5b) consists of intransitive verbal expressions conveying the physical result of impact, the impact itself - the verbs 'thrilled' etc. - being then reduced to a participial apposition to the subject, as in 'I staggered back, thrilled with chilly horror' (1.41) or 'I shuddered, seized with horror' (1.134). One further field (LF6: 'vision', 'discovery', etc.) identifies apparitions, memory, anticipation or revelation as straightforward causes of that impact. Table 1 summarizes the lexical fields that exist in the corpus; some terms enter into more than one field; the nodal field chosen is given in bold type:

Table 1: Lexical fields in the corpus

LF1 EMOTION ['horror', 'terror', 'pity', 'amazement', 'astonishment', 'despair', 'disgust']

LF2 ATTRIBUTES OF LF1 ['deadly', 'chilly', 'secret', 'black', 'sudden'] LF3 EXPERIENCER ['I', 'my soul', 'my blood'; 'he', 'the landlord'; 'we'; 'they', 'my comrades', 'the/our crew'; 'the phantom']

LF4 PERCEPTION ['beheld', 'gazed', 'discovery', 'stared', 'perceiving', 'hear',

'saw', 'looking', 'vision'] LF5 CONFRONTATION:

(a) Transitive verbs of physical impact ['thrilled', 'left him thrilled', 'seized', 'chilled', 'struck', 'pierced', 'fill', 'made (my blood) congeal']

(b) Intransitive verbs conveying the result of impact ['staggered back', 'trembled', 'shuddered', 'was fixed to the ground', 'was standing like a statue']

LF6 CAUSES OF PHYSICAL IMPACT ['apprehension (of X)', 'deeds (remembered)', 'discovery', 'phantom', 'vision']

One peculiarity of this language is its flair for binaries. The lexicon exhibits a clear tendency towards pairing emotion terms; 'horror and pity' (2.106) echoes the Aristotelian 'pity and fear', while the combination of 'horror' with 'amazement' (1.65, 2.153) or 'astonishment' (1.78) follows Edmund Burke's definition of the Sublime in his classic Philosophical enquiry (1757):

The passion caused by the great and sublime in nature, when those causes operate most powerfully, is Astonishment; and astonishment is that state of the soul, in which all its motions are suspended, with some degree of horror.11

These and other synonymic binaries ('horror and black despair', 1.103, 'disgust and horror', 2.182) yield a degree of redundancy which will be of interest later. Syntactic parallelism obeys the same principle (in the examples the two members of the binary are underlined):

1.166 I was struck with surprise when I beheld Baron T - before me, and I cannot but confess that I was seized with horror, when I saw him 2.43 I made another discovery which thrilled my soul with horror: Perceiving traces of blood on the pillow, I was seized with a sudden terror

In all, we can say that linguistic units in this novel are not generally treated as autonomous items but primarily as members of specific groupings - of fields. We will now see that this same principle affects grammatical construction.

See Burke 1757 (II.i.), p. 53. My concern being here with the significance of a formulaic pattern in this one novel, leading to a definition of the word 'horror' from the corpus itself, it would not serve my purposes to let the results of my analysis be explained by prior definitions. For some comments on the semantics of 'horror' and 'terror' in the light of such concepts as Burke's Sublime see, e.g., Botting 2009, Hughes 2013.

4. The syntactic field

A second element which contributes to the 'family air' evinced by the twenty-three tokens of this pattern is syntax. If every instance in the corpus is viewed as a syntagma consisting of collocations that bind up to six lexical fields, four types of syntactic structure emerge:12

Type 1

experiencer + passive verb ofphysical impact + force/agent ['with' + (adj.) + emotion term]

my soul they

the crew my comrades

seized with





chilly horror

black amazement


In these passive constructions 'horror' and its functional synonyms designate a hostile or threatening force personified to some degree as an agent which thrills, strikes or seizes one.13 Corresponding to them we have active constructions in which 'horror' has rather an instrumental function:

Type 2

force/agent + active verb ofphysical impact + experiencer + instrument ['with' + (adj.) + emotion term]

the vision thrilled the apprehension [of X]

deeds which filled

the phantom left

a discovery which made

my soul

my blood

with deadly horror

thrilled congeal

These in turn are transitive versions of Type 3, which 'contains' Type 1 as a reduced apposition to the subject:

I have made ready use of such commonly available linguistic categories (experiencer, agent, and instrument) as appear in Downing and Locke (2002), with no partisan intent. Systemic functional terms (senser, phenomenon, process, etc. in middle and effective clauses: Matthiessen and Halliday 1997), terms from cognitive linguistics (agent, instrument, theme, etc.: Langacker 1991), or the earlier terminology of case grammar (Fillmore 1968) could be used to characterize the constructions.

The personification of horror becomes even more noticeable in other syntactic formations, such as 'we were seized by the chilly hand of horror' (1.136), to be studied in a separate article.

Type 3

experiencer + verb of bodily reaction + past part. ofphysical impact + agent ['with' + (adj.) + emotion term]

I staggered back, thrilled with chilly horror


was standing like a statue, horror and pity

shuddered, seized astonishment

Our crew was fixed to the ground,

Whether the active or the passive voice is used here, fundamentally the same idea is conveyed, and there is little to choose between 'I was thrilled with horror' (Type 1), 'X thrilled my soul with horror' (Type 2) and 'I trembled, thrilled with horror' (Type 3). Lastly, we have statements of perception which echo Type 2 from the experiencer's viewpoint:

Type 4

experiencer + verb ofperception + direct object + manner adjunct ['with' + (adj.) + emotion term]

we beheld X with chilly horror

I could not hear without disgust and horror

saw him with secret

gazed at

Two variants give the verb of perception in the present participle while promoting the verb of physical impact (Type 1) or of bodily reaction (Type 3) to the status of a main verb:

Perceiving traces of blood on the pillow, I was seized with a sudden terror (2.43) The phantom ... trembled violently ..., looking around with chilly horror (1.92)

Subordination and apposition allow for combinations of these types; syntactic permutations of six lexical fields convey different thematizations of one key event: perception (Type 2 and Type 4) accompanies or elicits horror, the impact of which (Type 1) causes a physical reaction (Type 3). Though not identical the four types are functionally equivalent and constitute what may be called - on the analogy of lexical fields - the syntactic field of this formulaic pattern, that is, the set of interchangeable configurations the corpus exhibits.

5. Phonological fields

Whereas the standard definition of the epic formula relies on 'metrical conditions' and is therefore unsuitable to the analysis of literary prose, certain sound

effects which are commonly associated with poetry do seem relevant to this novel, and by the side of lexical and syntactic fields there is reason to posit the existence of phonological fields. The concept is tentative, and further research will be needed to bring it in line with the other two; but alliteration, for example, is much too frequent to be casual, as the following instances show: 'horror'-'beheld' (x5); 'seized'-'single' (1.134); 'mind'-'moment' (1.137); 'perceiving'-'traces'-'seized'-'sudden'-'terror' (2.43); 'standing'-'statue' (2.106); 'sung'-'obscene'-'songs', 'hear'-'horror' (2.182); 'full'-'fill' (2.145); 'being'-'famished'-'infernal'-'abode' (1.103); 'seized'-'horror'-'saw'-'him' (1.166). Alliteration establishes phonological bonds between words (among which the persistent 'horror'-'behold' link clinches the visual nature of horror). In alliteration we obtain the opposite of synonymy - not formal variants of one idea but different ideas brought together by a formal trait. A principle of equivalence prevails in both systems but whereas in the lexical field different words convey similar meanings, in the phonological field different meanings are made to carry the same sound.

Items from a lexical field can both substitute for and combine with each other. At first sight only combination seems logically possible in the phonological field, but some evidence suggests that replacement on the basis of sound-equivalence can nevertheless be found too: a peculiar use of what can only be called rhyme - but with a substitutive rather than combinatory function - generates a field by bringing together three largely interchangeable members of LF5a: 'thrill', 'chill', 'fill'. If in poetry rhyme is found typically in line-final position, in the present corpus these three terms appear eight times in simple past or in past participles (and once in a present participle) governing 'with horror'.14 A striking instance occurs in the following:

O that I had not to relate deeds more glaring and abominable! deeds which thrill my soul with anguish, and pierce my guilty heart with a thousand daggers, pointed by unutterable pangs of a polluted conscience: However, I promised you a sketch of my whole life, and will be as good as my word: Although I shall not be able to give you a full narrative of deeds which fill my soul with horror, yet I will go on as well as I can.


The underlined expressions constitute a clear case of redundant parallelism enriched by the use of 'thrill' and 'fill' in the same syntactic slot. They are instruc-

These are 'thrilled with horror' (1.41, 1.103, 2.43, 2.106, 2.153), 'thrilling my soul with horror' (1.153), 'chilled with horror' (1.24, 1.155), and 'fill with horror' (2.145). Cf. 'filled with shame and terror' (1.28), 'filled him with terror' (1.151), and 'filled my soul with terror' (2.197). 'Bliss', 'apprehensions', 'bodings', 'despair' may also replace 'horror' in these constructions.

tive in another way: 'horror' and 'anguish' are in these clauses equivalent not only because of syntactic and semantic considerations but also because they occupy a two-syllable slot with initial stress in what turns out to be a trochaic tetrameter. The recurrence of such phrases as 'thrilled with chilly horror' imparts a falling rhythm - fundamentally trochaic, with occasional dactylic feet -to the language that conveys the experience. Other examples are:

I was seized with horror (1.51, 1.166)

struck with horror and amazement (1.65)

seized with horror and astonishment (1.78)

looking around with chilly horror (1.92)

thrilled my soul with horror, and black despair (1.103)

we beheld with chilly horror an human figure hovering on the wall (2.11)

(Mark the persistent h-alliteration in this last example). Of course, prose cannot sustain such obvious cadences for long; but subtler variations on this and other rhythms can be detected elsewhere in the corpus. It would appear that something like the 'metrical conditions' postulated by Parry for the formula in oral epic poetry remains an ancillary concern in the organization of certain kinds of prose formulaic discourse.

6. Connotations

It has been shown that fields play a key role in the construction of formulaic patterns; but what can be concluded from their proliferation? Anthropology furnishes an instructive analogy in ritual theory. Ritual is not a distinct category of action but a modification on existing categories; in the wake of Huxley (1914) biologists, anthropologists and ethologists have been stressing the concept of ritualiza-tion as one key to our understanding of ritual behaviour;15 in Catherine Bell's (1992: 90) formulation, '[t]he significance of ritual behaviour lies not in being an entirely separate way of acting but in how such activities constitute themselves as different and in contrast to other activities.' Customary ritual strategies - stereotyping, periodization, stylization, iteration, fixity, as well as distortion or inversion of the rules prevailing in ordinary reality or in ordinary discourse - all presuppose prior non-ritual activities; it follows that in order to succeed as ritual the performance must bring the participants' and audience's attention to its own differentiation techniques and make these visible by adding to them what anthropologist Siegfried Nadel (1954: 99) called 'some conspicuous regularity not accounted for by the professed aims of the actions'.16

15 On ritualization see also Nadel 1954, Gluckman 1962, Huxley 1966.

A more abstract way of formulating this would be found in Erving Goffman's (1974: 43-4)

In a sense, of course, it can be said that for a century the literary text has been known to practice nothing else: such critical concepts as defamiliarization, intertextuality or trace11 presuppose a view of the literary text as both leaning upon and transforming prior texts. But different turns are applied to this screw whenever a literary narrative (or genre) seeks to differentiate itself from some existing narrative practice, and not all such turns confer ceremoniousness to a piece of writing; Teuthold's formulaic patterns exhibit features which are the narrative counterpart of those found in ritual practices - a surplus of organization not demanded by the strictures of the plot, an overpatterning which imparts to text 'conspicuous regularity'.18

One of the ways in which this literary 'ritualization' bestows visibility upon language is through foregrounding secondary ranges in its semantic system, and this is achieved by inscribing words, structures or even sounds within fields. Making a linguistic item into a member of a class reduces its autonomy and overloads it instead with connotative meanings derived primarily from its inclusion in the class. The connotations of 'horror' in The Necromancer are of two kinds: those inherited through historical development and cultural contextuali-zation, and those that result from its inscription within the various fields in the novel. The first group includes such items listed in the Oxford English Dictionary as a shuddering with terror and repugnance, which combines a) emotion terms roughly synonymous with 'horror' and b) words for physical aspects or reactions. To (a) belong the following OED senses: aversion and dread, loathing and fear, awe or reverent fear, the feeling excited by something shocking or frightful, etc.; to (b) belong roughness, ruggedness; nauseousness; thrill, shivering; and the obsolete ruffling, rippling of a surface.

Our corpus suggests much more. To begin with, horror is presented as arising from the individual's (predominantly visual) perception of a numinous figure, place, sound or narrative. Secondly, horror is disaggregated into (a) a lurking force acting upon the individual, (b) its intense physical impact, and (c) the concomitant bodily reaction this impact elicits. The frequent use of the passive voice and/or of past participles stresses the individual's quiescence in the face of horror and thereby the arresting nature of the experience, and parallels the characters' or narrators' inability to communicate their emotion.19 'Unable to utter a single word' (1.134) is the counterpart of 'fixed to the ground' (1.78) or

notion of keying as a strategy whereby 'a given activity, one already meaningful in terms of

some primary framework, is transformed into something patterned on this activity but seen

by the participants to be something quite else.'

See, respectively, Shklovsky 1917, Kristeva 1968, and Derrida 1967.

See Aguirre 2007, pp. 115-20.

On this ancient rhetorical device, the inexpressibility topos, see Curtius 1948.

'standing before him like a statue ' (2.106) This, too, is the logic of the persistent association of horror with extreme or excessive coldness; 'chilly' (x4), 'chilled' (x2), 'congeal' (1.153) convey an experience also implied by 'seized', 'struck', 'pierced', 'amazement' and 'astonishment': horror freezes individuals or otherwise renders them unable to react; or, on the contrary, in its wake they experience involuntary physical reactions ('shuddered', 'staggered', 'trembled'). The common theme is loss of control in the face of the Numinous. From this survey of the system of connotations which the formulaic pattern establishes around the word 'horror' it can be inferred that physical arrest and/or privation by an overwhelming power is one dominant feature in the system.

7. Visibility.

Emphasis on the physical quality of the emotion has its own rationale: it is a convention of the genre that the inner must be given an outward expression, and thoughts and feelings will manifest themselves in setting, atmosphere or bodily stance. This reiterates two principles of composition that we have already come across: equivalence, and visibility. By the one, everything mirrors something else; by the other, meaning is displayed rather than just conveyed. The first relies on a rhetoric of binaries and demands resemblance and difference, substitution or redundancy, harmony, resonance or conflict as part of the organization of the text; the second privileges image over narrative, the explicit over the tacit, surface over depth.

Gothic has often been described as resorting to 'techniques of the surface',21 and though the term is susceptible of receiving a demeaning sense (as betokening the 'superficial')22 it remains a suitable descriptor if we do not think of the genre as sadly unable to plumb 'depths' - depths which, in any case, were not conceptually available in the late eighteenth century; nor is 'depth' the only contrary of 'surface'. The overpatterning of the word 'horror' plays down its denotative meaning, foregrounding instead its connotative system and thereby refashioning the word as a mantra: 'horror' ceases to convey subjective emotion and becomes expressive of some ultimate truth. To be sure, this is not the truth of our ostensive reality, rather it pertains to some domain the unexpected discovery of which elicits horror; it is disturbing enough as it entails the existence of another order of things lurking 'beside', 'behind', often 'beneath' the light of common day - an order which commands awe. But though such a revelation

All three expressions are formulaic inasmuch as they appear elsewhere in the novel: 'unable to utter a single word' (x3), 'fixed to the ground/floor/our seats/their places' (x6), 'like/as a statue' (x5).

See Sedgwick 1981, Napier 1987. See, e.g., Thompson (ed.) 1979, p. 4.

will pave the way for Freud's discovery, we are not to think of this other domain as - at least in the eighteenth century - anything like the Freudian unconscious but as the 'other-side' of our surface: it is not quite distinct from our own domain - it is a pleat or an undersurface, the seamy side rather than a depth.

Motivated by the conviction that something vital is being left unsaid in the mainstream novel which from Defoe to Austen seeks to etch the reality of the human condition, for the first time a literary genre sets out, almost programmatically, to bestow visibility on aspects of the real (whether of the mind or of the world) which are acknowledged to exist but for which no definite expressive apparatus is available as yet. The principle of equivalence provides a means: an unknown thing may be represented by a known one; the restriction imposed by the principle of visibility, however, is clear: things must be not so much 'translated' into another, 'deeper' semantic level as expanded or unfolded along a single horizontal plane of meaning - even if this plane is conceived of as corrugated, pleated or infolded upon itself. For this, not metaphors or symbols but metonyms are to be favoured, and the physicality of the body and environment is found to be the simplest metonymic correlative of psychic experience, granting it visible traits. Hence it is that horror is made to inhere in sight or sound, that the contemplation of the dreadful must necessarily seize, strike, thrill or chill the experiencer, and that his response to it must be of a bodily nature (shuddering, trembling, staggering, or standing like a statue). At every turn the horror must be visible to readers.

Needless to say, the Gothic novel is not the sole practitioner of this technique: the declamatory and effusive stance of sentimental fiction, Gothic drama23 and melodrama are guided by a similar bid for visibility. Much as the copious tears of Mackenzie's Man of feeling or Richardson's Pamela are added to and become the tangible expression of their sentiments - but do not 'substitute' for these -, and much as Mrs. Siddons moved her audiences with her 'rhetoric of the passions',24 so the 'inward sensations of horror' experienced by Teuthold's characters (page 22 of volume I) manifest themselves in visible -typically physical - ways. Nor is the bodily the only correlative of psychic experience; analysis of formulaic diction reveals that buildings in The Necromancer tend to be used as adjuncts to otherwise hard to define states of mind.25

8. Conclusions

In light of the above analysis, the connotative system of 'horror' in The Necromancer can be said to include the following features: 1) horror arises from per-

See Ranger (1991: 90-105).

See Clery (2004: 17).

For this see again my 'Tranquillity'

ception; 2) it is a hostile force which 3) exerts a physical impact and 4) elicits a bodily reaction; 5) it exhibits arresting power, 6) is characterised by inexpressi-bility, and 7) causes extreme coldness and 8) loss of control. No discrete elements in the formulaic pattern of 'horror' are stable - not the experiencer, the narrative voice, narrative strand or any structural arrangement, and certainly not the node-word, which can be replaced with any number of alternative terms from LF1;26 only the pattern itself abides - a textual constant independent of plot, character or narrator. 'I was seized with horror, when I saw him' (1.166) is not significant in itself, nor as the utterance of one given individual, but because it resonates to two dozen similar instances. In a narrative flow which has been deemed incoherent,27 the formulaic pattern actually operates as a narrative anchor for the reader. Because what counts are not the characters that move around in the novel but the affective set-ups they are drawn into, psychology (in our conventional sense) is fairly absent from such a presentation; not individuals but type-emotions are the focal point. It is noteworthy that the twenty-three sentences in the corpus are spoken by no less than seven narrators on different narrative levels. Admittedly most of them are reported by Lieutenant B - , his voice and those of Hellfried and Herrman being the only ones mediated solely by the frame narrator; but it is equally true that narrative voices themselves are heavily conventional and not truly distinct in this novel. Furthermore, the twenty-three tokens are governed by a number of grammatical subjects (first and third persons, singular and plural) with shifting referents; the seemingly motley lexical field (LF3) this shape is nevertheless instructive: just about anyone can be at one time or other prey to the stock experience of horror, and included among the sufferers is one whom we would conventionally expect to be the cause of the horror - the phantom (1.92).28 This means that, whether reporting their own or others' experiences, different speakers formulate in the same words what, in the final analysis, is one and the same emotional event routinely fallen upon. Such 'conspicuous regularity', to use Nadel's term, transposes personal concerns to a universal key and absorbs them into the strictures of an order which is independent of (when not inimical to) human order; this ritualiza-tion of narrative generates a haunting moment - both habitual and disquieting -in which horror, conceived as a physical force, again and again impinges in a violent yet ceremonious way upon individuals with a paralyzing effect.

If the analysis provided for 'horror' reflects the facts in the novel, we can fine-tune the definition of the formulaic pattern slightly as an open system of

26 It can also be combined with these; for an example of redundant association of 'horror' and 'terror' see 2.43 (on the possible distinction between these see references in note 11).

27 See, e.g., Sadleir 1927, p. 17, Thompson 2002, p. 40.

28 As he obviously is in other instances (1.155, 2.11, 2.153).

lexical, syntactic and phonological fields which is regularly employed under certain conditions to convey an essential idea (the 'conditions' remain to be identified, and so this remains a provisional definition.) A discourse that relies on such constructs constitutes an extended speech act that minutely but persistently arrests the reader. It has a sort of stroboscopic effect, creating discontinuities in the narrative flow and simultaneously establishing a different kind of continuity - one not based on temporal, causal, or logical sequence but on attractions set up athwart sequences, so that linear development29 is checked by a supra-linear network of resonances. Because through repetition, collocation, combination, alliteration, rhyme, assonance, rhythm, redundancy or parallelism - through metonymic strategies - it draws attention to the weft itself of text, formulaic discourse partakes of the poetic function of language in that it 'projects the principle of equivalence from the axis of selection onto the axis of combination', focuses on the message itself and promotes what Jakobson (1960: 71) called 'the palpability of signs'.

Through this relative playing down of the referential in favour of the poetic function of language (also manifest in Teuthold's frequent use of prosodic effects) the text destabilizes its expected linear configuration and invites a reading in which all occurrences of 'horror' appear as but versions of one and the same emotional event. The ritualizing and universalizing quality imparted to the word 'horror' in the novel is a direct outcome of the regular use of this formulaic pattern. The speech act the text performs is one of awed monstration, and this is congruent with a work one main purpose of which is to sway the readers' passions by confronting them with a larger-than-life reality. This usage of the poetic function of language strongly argues that some of this prose should be assessed in terms of criteria not commonly appealed to in the evaluation of prose fiction.

Several lines of enquiry suggest themselves for further exploration. Perhaps not just The Necromancer but the Gothic genre itself resorts to formulaic discourse to achieve visibility, both in an effort to differentiate itself from the conventional eighteenth-century novel and as a means to overpattern and 'ritualize' experience - to enhance and at the same time control or contain the horror. Another hypothesis is that - given that all literature overpatterns to some extent -we should expect types and shades of 'ritualization'; since in The Necromancer degrees of the formulaic vary from one page to another, the proof of this hypothesis must emerge from an analysis offormulaic density. Nor will the study broached in this article be complete without, first of all, an analysis of all remaining tokens of 'horror' in the novel that will appraise in detail the connota-tive system of the word; further research into the nature and role of phonological fields is also needed.

On the linear nature of discourse see Bolinger 1952 and Waugh 1980.

There remains, too, the observation that in The Necromancer, over and above the seemingly random, episodic or repetitive nature of its plot, textual coherence is ensured by recurrence and self-similarity; these are features of a fractal organization, and the analogy with mathematical fractals will be worth exploring.30 If it is conceded that the same compositional principles govern the arrangement of lexical and other units into fields and of fields into formulaic patterns, the next question to be asked is, are formulaic patterns autonomous, or will there be evidence, on some 'higher' textual level, of larger structures within which patterns combine in a similar manner? This and other questions will be taken up in subsequent work.31


Aguirre, Manuel. 1990. The closed space: Horror literature and western symbolism. Manchester:

Manchester University Press. Aguirre, Manuel. 2007. The thresholds of the tale: Liminality and the structure of fairytales.

Madrid: The Gateway Press. Aguirre, Manuel. 2013a. Gothic fiction and folk-narrative structure: The case of Mary Shelley's

Frankenstein. Gothic Studies 15(2). 1-18. Aguirre, Manuel. 2013b. A grammar of Gothic: Report on a research project on the forms of the Gothic genre. Romantic Textualities: Literature and Print Culture, 1780-1840, 21. (date of access: 27 July 2014). Aguirre, Manuel. 2014. Mary Robinson's 'The Haunted Beach' and the grammar of Gothic.

Neophilologus 98(2). 1-16. Aguirre, Manuel. 2015 (in press). 'The tranquillity of the mansion': Fields and formulaic diction

in a Gothic novel. Journal of Language, Literature and Culture 62(3). Bell, Catherine. 1992. Ritual theory, ritual practice. Oxford: Oxford University Press. Bolinger, Dwight L. 1952. Linear modification. PMLA 67(7). 1117-1144. Botting, Fred. 1996. Gothic. London: Routledge.

Botting, Fred. 2009. Horror. In Marie Mulvey-Roberts (ed.), The handbook of the Gothic (2nd

edn), 184-192. Houndmills: Palgrave Macmillan. Burke, Edmund. 1757 (1987). A philosophical enquiry into the origin of our ideas of the sublime

and beautiful, ed. J.T. Boulton. London: Blackwell. Clery, Emma J. 2000. Women's Gothic from Clara Reeve to Mary Shelley (2nd edn.). Horndon,

Tavistock: Northcote House. Curtius, Ernst R. 1948 (1979). European literature and the Latin Middle Ages, transl. W.R.Trask. London: Routledge & Kegan Paul.

The analogy is proposed in my 'Tranquillity', and will be pursued in coming articles. My thanks to the anonymous readers of SAP for helpful remarks. I am also grateful to Beatriz Sánchez for debate and insights; and to the staff of the British Library for their un-

failing help.

Derrida, Jacques. 1967. De la grammatologie. Paris: Éditions de Minuit.

Downing, Angela & Philip Locke. 2002. A university course in English grammar. London: Routledge.

Fillmore, Charles. 1968. The case for case. In Emmon Bach & Robert T. Harms (eds.), Universals in linguistic theory, 1-88. New York: Holt, Rinehart, and Winston.

Fincher, Max. 2013. Terror. In William Hughes, David Punter & Andrew Smith (eds.) The encyclopedia of the Gothic (2 vols.), 683-686. Oxford: Wiley-Blackwell.

Firth, John Rupert. 1957. A synopsis of linguistic theory, 1930-1955. In Studies in Linguistic Analysis, ed. by the Philological Society, 1-32. Oxford: Blackwell.

Foley, John Miles. 1988. The theory of oral composition: History and methodology. Blooming-ton: Indiana University Press.

Foley, John Miles. 1990. Traditional oral epic: The Odyssey, Beowulf, and the Serbo-Croatian return song. Berkeley: University of California Press.

Fry, Donald K. 1967. Old English formulas and systems. English Studies 48. 193-204.

Gluckman, Max. 1962. Les rites de passage. Essays on the ritual of social relations. Manchester: Manchester University Press.

Goffman, Erving. 1974. Frame analysis: An essay on the organization of experience. Boston: Northeastern University Press.

Hall, Daniel. 2000. The Gothic tide: Schauerroman and Gothic novel in the late eighteenth century. In Susanne Stark (ed.), The novel in Anglo-German context: Cultural crosscurrents and affinities, 51-60. Amsterdam: Rodopi.

Hoey, Michael. 1991. Patterns of lexis in text. Oxford: Oxford University Press.

Hoey, Michael. 2000. A world beyond collocation: New perspectives on vocabulary teaching. In Michael Lewis (ed.), Teaching collocation: Further developments in the lexical approach, 224-245. Hove: Language Teaching Publications.

Hughes, William. 2013. Historical dictionary of Gothic literature. Lanham/Toronto/Plymouth: The Scarecrow Press.

Huxley, Julian. 1914. The courtship-habits of the great crested grebe (Podiceps cristatus); with an addition to the theory of sexual selection. In Proceedings of the Zoological Society of London 35. 491-562.

Huxley, Julian. 1966. Ritualization of behaviour in animals and man. In International study of the main trends of research in the sciences of man. UNESCO archives SS/41/3.244.1/h/13 (date of access: 6 April 2014).

Jakobson, Roman. 1960 (1987). Linguistics and poetics. In Krystyna Pomorska & Stephen Rudy (eds.), Language in literature, 62-94. Cambridge, Mass.: The Belknap Press.

Kahlert, Karl Friedrich (a.k.a. Lorenz Flammenberg). 1792. Der Geisterbanner. Eine Wundergeschichte aus mündlichen und schriftlichen Traditionen. Hohenzollern: Johann Baptist Wallishausser. (date of access: 28 February 2014).

Kristeva, Julia. 1968. Problèmes de la structuration du texte. In Roland Barthes, Jacques Dérrida & Michél Foucault (eds.) Théorie d'ensemble, 297-317. Paris: Seuil.

Langacker, Ronald W. 1987-91. Foundations of cognitive grammar (2 vols.). Stanford, Cal.: Stanford University Press.

Lord, Albert B. 1960. The singer of tales. Cambridge, Mass: Harvard University Press.

Louw, Bill. 1993. Irony in the text or insincerity in the writer? The diagnostic potential of semantic prosodies. In Mona Baker, Gill Francis, & Elena Tognini-Bonelli (eds.), Text and technology, 157-176. Amsterdam: John Benjamins.

Matthiessen, Christian & M. A. K. Halliday. 1997. Systemic functional grammar: A first step into the theory. Beijing: Higher Education Press.

Murnane, Barry. 2010. Uncanny translations, uncanny productivity: Walpole, Schiller and Kahlert. In Stephanie Stockhorst (ed.) Cultural transfer through translation: The circulation of enlightened thought in Europe by means of translation, 141-165. Amsterdam: Rodopi.

Nadel, Siegfried F. 1954 (1970). Nupe religion: Traditional beliefs and the influence of Islam in a West African chiefdom. London: Routledge & Kegan Paul.

Napier, Elizabeth R. 1987. The failure of Gothic: Problems of disjunction in an eighteenth-century literary form. Oxford: Oxford University Press.

Parry, Milman. 1930. Studies in the epic technique of oral verse-making. I: Homer and Homeric style. Harvard Studies in Classical Philology 41. 73-147.

Punter, David. 1980. The literature of terror: A history of Gothic fictions from 1765 to the present day. London: Longmans.

Ranger, Paul. 1991. 'Terror and pity reign in every breast': Gothic drama in the London patent theatres, 1750-1820. London: The Society for Theatre Research.

Sadleir, Michael. 1927. The Northanger novels: A footnote to Jane Austen. London: Oxford University Press.

Sedgwick, Eve Kosofsky. 1981. The character in the veil: Imagery of the surface in the Gothic novel. PMLA 96. 255-70.

Shklovsky, Viktor. 1917 (1965). Art as technique. In Russian formalist criticism: Four essays, 324. transl. Lee T. Lemon and Marion J. Reis. Lincoln, Nebraska: University of Nebraska Press.

Sinclair, John. 1991. Corpus, concordance, collocation. Oxford: Oxford University Press.

Sinclair, John. 2004. Trust the text: Language, corpus and discourse. London: Routledge.

Teuthold, Peter (transl.). 1794. The necromancer: or The tale of the Black Forest: Founded on facts. Translated from the German of Lawrence Flammenberg. In two volumes. London: Minerva-Press.

Thompson, Gary Richard (ed.). 1979. Introduction to Romantic Gothic tales, 1790-1840. New York: Harper and Row.

Thomson, Douglass H. 2002. Jane Austen and the Northanger novelists. In D. H. Thomson, Jack G. Voller & Frederick S. Frank (eds.) Gothic writers: A critical and bibliographical guide, 33-47. Westport, CT: Greenwood Press.

Waugh, Linda R. 1980. The poetic function in the theory of Roman Jakobson. Poetics Today 2(1a). 57-82.

Wray, Alison. 2005. Formulaic language and the lexicon. Cambridge: Cambridge University Press.