One of the most frequently asked questions, but one that is most difficult to answer, is: Can the date of production of a document be determined? The reason that this request is so common is that often documents are created at some time other than their purported date to justify some particular claim. For example, an agreement may be drafted but given a date that is years in the past to justify some financial gain or other.
It perhaps should be made clear at this point that dating a document does not refer to determining the actual day on which a document was produced! Absolute dating is often an approximate term, for example finding evidence that a document was produced after a particular date or period ('the questioned document was produced after 1950 because…'). Relative dating is where a document's time of production can be established in relation to some other document's production ('the questioned page was written before/after some other page').
The possibility of dating when a document was produced will be significantly influenced by the nature of the document concerned. In this chapter, dating options will be considered for a number of document production methods, starting with a simple handwritten document.
The dating of handwriting (as opposed to the dating of ink on paper) is a possibility that depends upon the time factor in a particular case. Handwriting changes in a given writer are most rapid in the young (typically those under the age of about 20), the elderly (extremely variable, but typically by about 70 years of age handwriting may start to deteriorate), and in those with certain medical conditions that affect handwriting production (see Chapter 2). The likelihood of being able to date handwriting production will be considerably affected by the time gap involved. For example, say the time gap in question was five years with a document having been written five years before (or after) it purportedly was written. If a document was alleged to have been written when the writer was aged either 13 or 18 then it may well be possible to reach a conclusion providing suitable samples are available. Such a five-year gap when comparing handwriting from a person aged 40 and when they were aged 35, however, is much less likely to yield any notable differences as handwriting at these ages is generally stable and unlikely to have demonstrably changed. Then again, a five-year gap between 75 and 80 years of age in a writer may well reveal a notable deterioration, particularly if associated with medical conditions such as the onset of Parkinson's disease.
If instead of five years the gap is one year or six months or one month, then clearly as the gap decreases the less likely it will be that there is any evidence of change, even in younger and older people whose handwriting may alter more rapidly. Conversely, if the time gap at issue were to increase from 5 years to 10, 20 or more years, then the chances of finding differences will increase across all age groups.
The most important factor in such cases is obtaining reliable handwriting samples, as the date on which they were written is crucial to being able to carry out the relevant examination. Finding specimen documents written in the past, especially several years previously, and which bear dates or whose date of production can be approximated (for example in school books or diaries) is by no means readily achieved in many cases. Also, when dealing with medical conditions, the nature of the illness and the taking of any medication (which may lead to short-term improvements in handwriting movement control) are relevant factors to consider, and it may be that expert medical input is needed to confirm the impact that the condition has on a person's ability to write.
As described in Chapter 2, the way in which a person writes is influenced by their educational experiences, which in turn vary in time and place with different styles being taught in different countries and with more styles replacing older ones. This creates the possibility that allegedly historical documents may be written in a style that was not used at that time.
Dating when an ink-written document was created has been extensively studied. For a review of much of the research, Brunelle's book listed in Further Reading in the Preface gives a lot more detail than can be given here.
Perhaps the simplest approach is to determine the nature of the components present in an ink in order to determine whether those components were available at the relevant time. For instance, if a document is dated 1928 but was written with a ballpoint pen, then that would not be possible since the ballpoint pen was not invented and available until a few years later. Likewise, as the components of ink have become more complex over time, if it is known when a certain component was first used then this gives an earliest date of availability for inks incorporating it. The difficulty with such an approach is that much of the information relating to the introduction of different ink components is commercially sensitive and is not easily discovered (Weyermann et al., 2012). It is possible instead to collect samples of inks and analyse them and thereby create a database of inks and link this to information on when they were first available. This resource can then be used to compare questioned inks against the database in an attempt to determine its (approximate) date of production or to at least determine when it could not have been produced. Indeed, for a time various tags (such as rare earth elements not normally found in inks) were added to inks at different times to make dating them more feasible, but this had modest uptake with the manufacturers.
There are in principle two types of ink dating question that can be asked: (i) When was the ink placed on the document? (ii) When was the ink of this entry placed on the document in relation to other entries (usually on the same document)? In other words, absolute dating and relative dating of ink, respectively. There are two important types of confounding factor in ink dating problems that centre on:
Despite the difficulties, much research has been carried out addressing the ink dating problem, particularly the relative dating since, in most cases, any confounding factors should apply equally to all parts of a document (assuming it was created with the same materials and stored in similar conditions).
The drying of ink on the page is perhaps an obvious theoretical starting point for the dating of ink. Some components of ink evaporate relatively quickly or else diffuse into the paper, whereas other components will be absorbed and adsorbed by the paper substrate (Weyermann et al., 2012) (see also Box 10.1). As a consequence of the various types of process involved, the drying of ink on paper is a complex process (Cantu, 2012).
Gel pen inks (which are water-based but which contain a mix of other components such as glycerol, ethylene glycol and diethylene glycol) were studied by Li and colleagues using gas chromatography (Li et al., 2014) showing that some volatile components of the inks evaporate rapidly over the first days and that artificial ageing causes loss of the volatile components in a matter of hours.
The drying process not only leads to the loss of the more volatile components of the ink, it is also associated with changes in the physical properties of other components, particularly the resins present, which may tend to harden and thus become more difficult to extract from the document. This leads to the idea that the longer the ink has been on the document, the more difficult it is to extract using a solvent. An extension of this idea is that at some point the changes to the ink will reach an end point and stop. Hence, if this end point could be determined for the ink on a particular document, the extent of the ink's time passage towards that end point would be an indication of how long the ink had been on the document. From this comes the idea of what amounts to the artificial ageing of the document. By warming the document under controlled conditions the natural process of ink drying can be sped up to reach its end point. If the extractability of the ink ‘now’ (before the document is artificially aged) is determined (perhaps giving a value of X) and this is compared to the extractability after artificial ageing (giving a value Y), then X should be greater than Y and by an amount that indicates how far short the ink is of reaching its end point.
Using the artificial ageing approach requires some knowledge of the rate of change of extractability of ink from a document. Many studies have found that the process is not linear but rather appears to be exponential. In other words, the rate of change in extractability is rapid to begin with (essentially meaning that many of the changes due to drying are relatively rapid over days and weeks) and that thereafter the changes in extractability are much slower over subsequent months and years, with the proviso that, for a given ink and document, the rates of change will depend on the confounding factors noted above (Weyermann et al., 2012).
The many methods of ink comparison have often been employed to address the ink dating issue as well as the more common problem of distinguishing between different inks (see Chapter 7 and Box 10.2). The important point is that, as for any technique, the methods used should be thoroughly tested and validated before being used in the forensic arena and that uncertainties caused by the potential confounding factors of materials used and storage conditions must be factored in as potential sources of error. Just as with other document dating problems, the time gaps that are relevant to a particular case will have a major impact on the usefulness (and especially the reliability) of evidence from ink dating methods.
Paper is a complex material with many different organic and inorganic components (see Chapter 6). The (approximate) date of the introduction of some components into paper manufacturing may be known and can provide information on the earliest date on which such papers could have been available. A well-known example was in the case of the forged Hitler diaries in the 1980s. Optical brighteners were first introduced into the paper industry in the late 1940s, and yet the diaries, allegedly describing events before his death in 1945, were in part written on paper containing such brighteners. Indeed, not only the paper but also other components of the binding of the diaries were shown to have not been available until years later.
Paper manufacturers generally keep records of what materials they use to make their papers and other information relating to production, such as the dates on which watermarks are introduced. Such background information can be a way of determining when a piece of paper was made. The dates on which different watermarks were introduced by a paper manufacturer have enabled the date of production of a will to be determined (Allen & Rimmer, 1988). Similarly, the use of a particular set of pulp fibres in a paper's manufacture again led to the dating of its production by cross-checking with the records at the paper mill (Totty et al., 1987).
Despite the occasional success in dating paper and related products, such successes are rare because most materials used in document production are widely available and remain unchanged over long periods of time. New products, however, do provide potential for dating documents and, together with manufacturer's records, there may be a chance of a successful forensic investigation. That said, the vast majority of paper is featureless (for example having no watermark) so identifying the manufacturer in the first place is extremely difficult.
The fonts used in typewriters changed over time as new typefaces were introduced. The rate of newly conceived styles increased rapidly once computers became more commonplace. But the fact that typestyles have a date when they first were available again provides a possible mechanism for dating documents.
As with other dating methods, the longer the gap between the purported date of production and the actual date of production (assuming that indeed the document is fraudulent), the better the chance that there will be something wrong about the fake document. The date on which a particular typeface became available is likely to be difficult to establish with any precision and indeed its availability may vary from place to place since many products may be launched in one country before becoming available elsewhere, for example.
At a more general level, different technologies have been introduced over time, although again they will not have a precise launch date but rather became more and more widely available. For example, inkjet technology was introduced during the late 1970s, so a document dated 1965 that was inkjet printed is unlikely to be genuine. Further, inkjet technology has improved over time with, for example, ever smaller droplet sizes becoming available, so if the drop size can be calculated (with reference to the dot size on the paper) it may be possible to show that a document bearing a date of, say, 1985 could not have been produced at that time because the dot sizes are too small; in other words, the quality of the technology used to produce the suspect document was not available at the time when it was purportedly created.
Printing devices will tend to become damaged with use (see Chapter 4). Any change over time provides the potential for being able to demonstrate a chronological sequence of changes in documents produced on a device over time. For example, a brand new typewriter or computer printer will probably produce well-aligned, clear typed output, but over time with much use characters may misalign or become damaged or various components may malfunction such that the typed product is no longer ‘perfect’. Documents produced during this period of deterioration are a record of the changes and if they are dated (or their date of production is known) then this becomes a reference chronology against which a suspect document can be compared. A case example that shows this made use of the fact that the gradual break up of a single letter (lower case k) on a printwheel typing element occurred over a period of months and the chronology that could be established enabled the suspect document's date of production to be established (or more precisely a period of time during which it must have been produced) and this did not correspond to the date on the suspect document (Hardcastle, 1986).
As noted in Chapter 4, some changes that occur in mechanical devices such as printers and copiers may be transient. For example, servicing may improve a defective copier or replacing a printer cartridge may change the appearance of a computer printer's output. However, as defect marks on a photocopier, for example, accumulate over time, they will create a chronology that can in principle be used to date when a questioned document was produced on that copier during this period of time. Such factors must be taken into account when assessing the evidence in relation to establishing reliable chronological sequences of documents.
The examination of pages in such multi-page documents is not perhaps strictly speaking a dating issue as it is normally considered, but rather it is concerned with showing when pages were created in relation to one another.
A suspect document can either bear impressions of handwriting that come from some other document (A) or the handwriting from a suspect document can appear as impressions on some other document (B). Which of these two situations occurs will be dependent upon the presence of sheets of paper in a stack or pad at the time. If either of the other documents (A or B or both) bears a true date (in other words the written date is the actual date that the document was written), then it might be possible to show whether the suspect document was written before or after one or other of A or B by following the approach described for sequencing handwriting and impressions of handwriting in Section 9.1.3 of Chapter 9 to determine whether the handwriting or the impressions of handwriting were created first on the page.
This complicated situation can be illustrated with a short example. An invoice is written out to cover up a financial fraud. The invoice is from a pad of invoices and the suspect invoice dated 1 March is number 20 in the pad. Examination of invoice number 21 is dated 1 July in the same year (its true date of being written) and it shows impressions of handwriting from the writing on invoice number 20 and these impressions overlap in some places with the actual handwriting on the invoice number 21. An electrostatic examination of invoice number 21 indicates that the handwriting on it was written before the impressions were created (from the handwriting invoice number 20). This would then show that invoice number 20 was written at some time after invoice number 21 was written out, in other words at some point after 1 July and not on 1 March.
If an upper document is written on a pad resting directly on a sheet below that already bears handwriting, then the pressure applied during the production of the handwriting on the upper document may cause some of the ink on the page beneath to be transferred onto the reverse of the upper document. The areas of ink transfer will correspond to points when the writing on the upper sheet crossed the handwriting on the lower sheet. If such ink transfer occurs and can be demonstrated, it shows that the upper document was written at some time after the lower document was written out.
Some documents consist of many pages and are written on pages that form a stack or (if bound) a pad of paper. In such cases, a series of pages are supposedly written in sequence such that page one was written out whilst resting on page two and then put to one side when the page was full, then page two was written on whilst resting on page three and so on down the stack. An example might be simply writing a long letter over several pages on lined paper from a typical A4 pad. Another example would be taking down a written record of some kind on pages in a stack, and over the years the prime example of this has been police notes of interview. If at some point the handwritten record is changed for some reason and a sheet of paper discarded and re-written, then this will be apparent from an examination of the impressions of handwriting on the relevant sheets. For example, if a mistake is made half way down page four and this page is discarded and a new page four is written on the next available sheet in the stack, then examination for impressions of handwriting on this new page four may show the content of the discarded page four.
The order in which the many different media that can be used to produce a document were placed on the document may be important to establish (for simplicity, all media will be referred to as using inks in this section). Given the many ways of creating a document (ranging from use of pens to typewriters and printers and printing to stampmarks) it is not surprising that the methods to determine their sequence of placement on a document vary across those described in Chapter 7, although some techniques have been developed to specifically address the sequencing problem in some cases.
The interaction between the first and second ink will depend on two key factors. First is the nature of the inks, whether they are water based or oil based (Ozbek et al., 2014) or some other medium (such as toner). Second, the time gap between when the first ink is placed on the document and the second ink is then used will affect the extent to which the first ink has had an opportunity to dry. In terms of the forensic examination, a third parameter is the time lapse between writing with the inks and the examination being made. There are, therefore, many potential problems when examining such cases and great caution is needed when interpreting the observations.
Perhaps the most frequently encountered sequence problem is where two ballpoint pen ink lines cross (for example, where a body of text intersects an associated signature and the signatory alleges that some of the text has been added after the document was signed). A microscopic examination of the points where the ink lines intersect will rarely give a clear indication as to which line is first, and particular care needs to be taken when the two ink lines are of different intensity – as the darker ink will appear to be written over the lighter ink but this may be an optical illusion – so this observation should be treated with extreme caution.
A method that has been developed specifically for this type of problem involves the use of a high gloss card that is placed over the intersecting ink lines and heavy pressure applied, which has the effect of lifting some of the ink from the page onto the card (Mathyer & Pfister, 1984).
The interaction between the liquid ink from a stamp pad and other media, such as pen inks and toner, will depend to some extent on the amount of mixing between the first and second inks. Raman spectroscopy and high performance thin layer chromatography have been used with some success to determine the sequences in some instances (Raza & Saha, 2013).
The molten powder-like nature of toner from a photocopier or laser printer has the effect of the toner sitting on the surface of the paper with very little of the material absorbed into the paper surface. One method to examine the intersection is by using microscopy to look at the sheen and reflection of the toner, for example, and the effect of the ink on these properties (Saini et al., 2009). Caution may be needed in some instances, however, as when a liquid ink, such as from a gel pen, is applied before or after toner, the ink tends to seep through, giving the appearance that the toner was applied after the ink (Montani et al., 2012). Rather than looking at the inks themselves, it is possible to examine the micro-topography of the paper surface where the groove of the ink line and the toner material intersect using laser profilometry (a method of looking at the microscopic ‘hills and valleys’ present in a small area of the paper surface) (Montani et al., 2012). This method produces good results providing the paper surface is not too rough – which has the effect of making the hills difficult to distinguish from the valleys.
There are many other possible ways in which a document can be dated depending on its nature.
Backdating a document, particularly if it is many years later, is fraught with problems since it may be that the relevant batches of document are no longer available or the fact that telephone numbers have changed in the meantime is missed, or any of the other pitfalls described in this chapter have occurred. Such issues have been a significant component of the means by which a number of fraudulent documents have been exposed over the years.
In a slightly different context, fraudulent artwork shares many of the kinds of problems associated with documents, such as acquiring the appropriate materials relevant to a particular time when genuine art was produced by a given artist, and that is before the art is to be created in the correct and convincing manner expected of the artist concerned.
The method of dating a document will vary depending on which aspect(s) of the case were relevant. Thus it is not possible to generalise how case notes should be made or how the findings should be reported. However, dating is generally done either by an absolute process (such as showing when some paper was made by reference to manufacturing records, for example) or by a relative process (such as showing that a writer's signature deteriorated over time due to illness). In both situations, it is necessary to place the questioned document into a chronology that has been established by relevant information and thus a description of that chronology and the means for establishing it must form a part of the report. Then the placement of the disputed document into that chronology can be described so that the reader can understand how the document was dated.