Introduction

The following outlines AECOM’s general methods employed for the investigation of construction impacts to archaeological resources within the I-95 Girard Avenue Interchange Improvement Project area in Philadelphia, Pennsylvania. This document describes the broad types of archaeological resources/deposits known or likely to exist within the project’s area of potential effects (APE), the specific methodologies employed during the investigation and documentation of these resources, the procedures and studies performed during subsequent laboratory analyses of recovered artifact assemblages, and the ways in which the information gathered for the project is being disseminated. Inasmuch as the archaeological portion of the I-95 Girard Avenue Interchange Improvement Project continues to address construction impacts and design changes as they arise, this summary also represents a planning document to guide ongoing investigations.

A programmatic agreement (PA) between the Federal Highway Administration (FHWA), the Pennsylvania Department of Transportation (PennDOT), and the Pennsylvania Historical and Museum Commission (PHMC) governs the archaeological investigations within the I-95 Girard Avenue Interchange Improvement Project area.¹ In general, the investigations undertaken to date involved a four-pronged approach, incorporating and integrating archival and historical background research, controlled excavation and archaeological documentation, laboratory analyses of recovered artifact assemblages, and analysis and reporting. All activities and studies completed in conjunction with these investigations complied fully with the requirements of the PHMC’s Guidelines for Archaeological Investigations in Pennsylvania.²

In addition to the general approaches outlined below, site-specific data-recovery methodologies were developed for each site requiring Phase III mitigation investigations. Typically, a research design was established during in-field meetings between AECOM personnel and representatives from PennDOT and the PHMC. The basic units of archaeological inquiry and resource provenience for the investigation were individual historical house lots/properties, as identified through georeferenced historical mapping of the project area.

Background Research

Intensive background and archival research were conducted during all stages of the project. This research was begun before the field efforts, conducted concurrent with fieldwork and artifact analyses, and continues in the present. The research focused on collecting detailed information about the historical development, use, and occupation of properties over time, as well as information related to the development of the surrounding community. Background research also explored Native American use and occupation of the APE and involved the gathering of comparative data from the Pennsylvania Archaeological Site Survey (PASS) files pertaining to comparably aged sites previously identified in Philadelphia and the surrounding region.

In regard to the historical occupation of the APE, AECOM historians took advantage of the wealth of existing historical background information for Philadelphia and previously documented historical resources. Background research was conducted using primary and secondary sources from a variety of online and physical repositories held by the City of Philadelphia and the Commonwealth of Pennsylvania, including the Philadelphia City Archives and the Pennsylvania State Historic Preservation Office. In addition, research was conducted at the Historical Society of Pennsylvania, the Free Library of Philadelphia, the Library Company of Philadelphia, the Frankford Historical Society, and the Presbyterian Historical Society.

Property/Parcel Research

Specific background research was conducted to establish comprehensive occupational histories for historical properties subjected to archaeological investigation and\or data-recovery documentation, and to collect information related to the individual households who occupied these sites over time. As property addresses change over time, the address of a parcel is not an adequate means of denoting a discrete plot of land for the purposes of archaeological interpretation. Therefore, analysis and research focused on collecting information related to tax parcel IDs (numbers corresponding to deed chains for discrete areas of land as recorded at the Philadelphia Department of Records). Each tax parcel ID was then linked to a list of historical addresses relevant to that lot of land, enabling the discussion of a property to encompass all of the relevant historical address incarnations and the records pertinent to a particular lot of ground. In subsequent reporting text and for ease of reference, most properties are referred to via the parcel address derived from the lot as described on early-twentieth-century Sanborn insurance maps.

AECOM staff established contact with local historians and descendants of project area residents to access any information they possessed relative to the development and history of the project area. Research conducted on parcels that did not contain archaeological features classified as high priority (see the Prioritization of Features section below for rating criteria) concentrated on primary sources such as historical maps, deeds, census records, city directory listings, building permits, newspapers, and vital records. For records like deeds, city directories, death records, and censuses, data about individuals such as their name, place of residence, profession, race, gender, place of birth, relationship to the head(s) of household, age, and cause of death were collected and entered into a Microsoft Access database. This information could then be queried in various ways to build profiles of individuals, parcels, or whole city blocks, enabling archaeologists to have contextual information about the residents in occupation at archaeologically investigated parcels, and to provide quantitative context and demographic insight into the project area—enabling both site- and neighborhood-level analyses and interpretations.

For parcels containing features identified as high priority (see Prioritization of Parcels below), intensive research was conducted utilizing historical maps, deeds, census records, city directory listings, building permits, newspapers, probate records, and vital records, as well as secondary sources. Gathered information was entered into the Access database and is presented in the site reports in tabular form. In addition, the research was synthesized into property histories detailing the development of the individual parcels and the lives of those who lived and worked on them. If available, these property histories were supplemented with family traditions and genealogical records, including photographs and letters.

Primary source records—including census returns, city directories, religious records, vital records, and military records—were mostly obtained from Ancestry.com, FamilySearch.org, and Fold3.com, as well as PhilaGeoHistory.org’s online databases. Selected newspapers were obtained from online databases at genealogybank.com, newspapers.com, fultonhistory.com, and chroniclingamerica.loc.gov, while microfilmed newspapers were accessed at the Free Library of Philadelphia. Deeds and other property records were accessed through the phila-records.com database or at the Philadelphia City Archives. Additional supportive research was conducted, as needed, at such repositories as the Historical Society of Pennsylvania, the Free Library of Philadelphia, the Library Company of Philadelphia, and the Presbyterian Historical Society.

Controlled Field Investigations

Cultural Resource Potential

Based on the findings of a prior Phase IA archaeological sensitivity study,³ it was determined that the project area exhibited an overall high degree of archaeological resource preservation. Moreover, the findings of the Phase IA study suggested that the project area contained a wide variety of specific resource types, including those associated with both pre-contact/Native American and historical periods of occupation. Anticipated resource types were categorized by their interpretive potential, as detailed below.

Historical Ground Surfaces

Historical ground surfaces are represented by intact or partially impacted and naturally occurring soil horizons that contain deposits of in situ pre-contact/Native American and/or historical artifact deposits. Individual soil strata that fall within this category include those identified as A, E, and B horizons, and that either represent the present-day ground surface or are sealed beneath fill deposits of variable depth and thickness. In addition, other deposits that may qualify as historical ground surfaces include secondary A/Ap horizons formed within fill deposits and artificial fill deposits utilized at times in the past as living surfaces. Artifacts contained within these soil horizons take the form of sheet midden deposits that can provide significant information related to site chronology; the presence, distribution, and organization of functionally discrete activity areas; and the use of space within a given pre-contact occupation or historical lot.

Near-Surface Features

Near-surface cultural features may be either pre-contact or historical in age and cultural affiliation and may encompass all manner of intrusive excavations contained within, or immediately below, historical ground surface soil horizons. Within the context of the I-95 Girard Avenue Interchange Improvement Project archaeological investigation, such features were defined as exhibiting a final depth less than 4 feet below the present ground surface. Examples of near-surface features include postholes, lined and unlined pits, small barrel privies, hearths, drains and utility trenches, and both human and animal burials. Shallow features such as these can provide information related to the use of space, the presence and location of prior structures and landscape elements, and the nature of refuse disposal within a given site. Artifact deposits contained within or found in association with these types of features can be informative about a range of domestic, socioeconomic, and health issues associated with past occupants of a given site.

Deep Features

As defined for the purposes of this investigation, deep cultural features were likely to be exclusively historical in age and consist of intrusive excavations extending to depths of more than 4 feet below the present ground surface. Features falling into this category were limited to a relatively small variety of forms, including large barrel privies (consisting of two or more vertically stacked barrels), wood-lined box privies, and brick-lined shaft features (wells, cisterns, and privies). Wood-lined barrel or box privies can extend to depths of between 6 and 8 feet below ground surface, while brick-lined shafts can reach depths of 15–25 feet below ground. Unlike the first two resource categories above, excavation of deep features required the use of OSHA-compliant sloping, benching, or shoring systems to allow members of the archaeological team safe access to artifact deposits.

Historical Structural Remains

Historical structural remains consist of the intact or truncated remnants of former residential, commercial, and industrial buildings, including foundation walls, basement floors, pier footings, and related elements. Structural remains were not normally considered to represent potentially significant resources in their own right. However, given that such features can provide critical information related to historical property identification, the allocation and use of space within specific lots, the functional and chronological interpretation of associated features, changing land-use patterns, and overall site developmental sequences, all exposed structural remnants were mapped, documented on standardized forms, identified with respect to buildings featured on historical maps of the project area and surrounding vicinity, and photographed. In addition, all identified foundations were assigned sequential structure numbers within each archaeological segment of the project area and were linked wherever possible to specific street addresses.

In many cases, historical foundational remains and associated cellar holes were not accessible for full excavation and documentation. However, in instances where such investigations could be performed safely, a sample of historical foundations was excavated to the extent possible and recorded. Likewise, in regard to structures not depicted on historical maps of the project area—and that could be associated with early settlement and development in Philadelphia—all reasonable efforts were made to excavate and document such physical remains as completely as possible. The project principal investigators made decisions regarding the selection of historical structures for more complete investigation on a case-by-case basis in consultation with PennDOT and PHMC archaeologists.

Field Methodologies

The primary goal of controlled field investigations was to recover a representative/comparative sample of artifacts from a given historical lot (including artifacts that could be used to more precisely date identified cultural deposits), to expose any intact features present, and to document, recover, analyze, and interpret intact artifact deposits contained within deposits and features.

Locations for the controlled field investigations were predicated on whether any construction impacts would affect an individual area. These impacts typically consisted of the relocation of existing streets; the construction of support piers, expanded abutments, and temporary access roads; the installation of rain gardens and other stormwater facilities; landscaping activities; the construction of new entrance/exit ramps; and the installation and removal of subsurface utilities. The removal of embankments and road surfaces were also considered to be impacts requiring archaeological investigation. Some of the tested areas were revisited several times as additional construction impacts were identified.

The specific type of controlled field investigation was dependent on several factors. As multiple and deep layers of modern fill capped most of the areas to be impacted, heavy machinery was necessary for removal. In general, this machinery was initially used to open test trenches with the purpose of ascertaining the presence or absence of historical ground surfaces or archaeological features. If the trenching encountered a buried historical ground surface, the fill layers were removed by machine and testing of these historical ground surfaces proceeded as described below. If no such surfaces were encountered, but historical features were exposed, testing of the near-surface and deep features proceeded as described below as well.

Ground Surfaces

Intact pre-contact or historical ground surfaces and soil horizons selected for data-recovery investigations were mitigated through the completion of a series of 3-x-3-foot or 5-x-5-foot excavation units (EUs). All EUs were excavated by discrete natural strata and in arbitrary 4-inch (about 10-centimeter) levels within undisturbed soil horizons. Excavations proceeded to a depth of at least 4 inches into sterile subsoil deposits. All excavated soil was screened through standard 0.25-inch hardware cloth, and all recovered artifacts were assigned a sequential field sample (FS) number. The FS number is linked to the precise location and stratigraphic context of those materials, which were retained in plastic bags labeled with all appropriate vertical and horizontal provenience information. Records of the unit excavations—including the Munsell color, texture, and depth of all soil horizons, along with the artifact content of each stratum—were recorded on standardized paper forms and maintained with the project files. Exposed soil profiles were recorded via hand-drawn maps and high-resolution digital photographs. Soil samples were retained for further study, where appropriate. The number and distribution of EUs necessary to effectively mitigate construction impacts was planned in consultation with PennDOT and PHMC archaeologists.

Near-Surface Features

Near-surface features were excavated by hand and initially bisected in order to reveal and document internal fill structure and stratigraphy. If bisection revealed the presence of no potentially significant artifact deposits, archaeological documentation terminated once this process was completed. If potentially significant artifact deposits were identified, full excavation (mitigation) of the feature was undertaken. Full excavation of select features was also executed if they were found to be unique in form and/or function, and in order to fully record the associated physical attributes of such examples.

During all excavations, feature fill was removed by observed internal strata; however, in some instances smaller arbitrary divisions were maintained within cultural deposits where such deposits exceeded 1 foot in depth and where the use of such finer divisions was likely to yield analytically significant data. All excavated matrices were dry-screened on site through 0.25-inch hardware cloth. Standardized paper field forms were used to record data relating to the depths, Munsell color and texture, and artifact content for each shaft fill stratum. Feature depositional profiles were fully recorded by means of hand-drawn maps and by fine-grained black-and-white and high-resolution digital photographs.

Soil samples were typically collected from artifact-bearing deposits within features. Samples averaged 2 liters in size and were floated at the conclusion of fieldwork, with subsamples from diagnostic features retained for possible future specialized studies.

Deep Features

Deep cultural feature documentation followed the same procedures and methods outlined for near-surface features. Standardized paper field forms were used to record data relating to the depths, Munsell color and texture, and artifact content for each fill stratum. Feature depositional profiles were fully recorded by means of hand-drawn maps and high-resolution digital photographs.

Deep features were bisected and excavated by observed stratigraphic deposits. For safety reasons, excavation of deep features (greater than 4 feet) was completed in vertical sections that did not exceed 4 feet in total depth, with data on internal stratigraphy progressively added to the feature profile and notes. In order to reach depths below 4 feet, OSHA-compliant sloping or shoring practices were utilized. If bisection did not reveal the presence of potentially significant artifact deposits, archaeological documentation was terminated. If potentially significant artifact deposits were identified, full excavation (mitigation) of the feature was undertaken.

While feature fill contexts were all primary archaeological contexts, in the sense that they had not been moved from the feature since their deposition, a further distinction for feature fills was established to highlight the relationship of the deposits to the actions which created them historically. For shafts, this included dividing privy deposits into primary privy deposits (night soils and remnant clean-out events from the use of these shafts as sanitary infrastructure) and secondary privy deposits (such as fills deposited during abandonment). Again, while both primary privy deposits and secondary privy deposits were stratigraphically classed as primary archaeological deposits relative to the site, their distinction relative to the feature and its formation provided an important analytical distinction.

All excavated primary privy deposits were screened through 0.25-inch hardware cloth and, where possible, samples of the sediment matrices were retained for flotation. The same was true of secondary feature fill deposits that were artifact-bearing and might be indicative of initial abandonment and/or discard related to the property’s residents. Such secondary feature fill deposits were important sources of dating information for when a feature was abandoned, making them important to establishing a temporal chronology. For secondary feature fill deposits comprised of bulk dumps of material (ash/cinder, construction/demolition rubble, and related fills), typically only a 25% sample was screened for artifact content. Field archaeologists collected temporally diagnostic artifacts within such bulk feature fills and retained them for further analysis. These bulk fills were sampled in the described manner, as their sediment matrix likely represented material imported from off site. Such material was often brought in specifically to infill an abandoned shaft. This historical occurrence prevents the reliable attribution of cultural material to the historical lot associated with that feature, limiting its interpretive potential to providing temporal information only.

Soil samples were typically collected from artifact-bearing deposits within features and retained for flotation and/or future specialized studies. Samples averaged 2 liters in size and were floated at the conclusion of fieldwork, with subsamples from diagnostic features retained for possible future specialized studies.

Modified Harris Matrix Context System

The archaeological landscape of Philadelphia has been rendered complex due to hundreds of years of human activity. This activity can be observed at a large scale, in things like the ever-changing aspects of the neighborhoods themselves due to development and industry, and at a small scale, in many of the shaft features that underwent various clean-out and depositional periods. Originally, a standard stratum/level system was utilized to order these various stratigraphic layers. In this standard archaeological system, stratum referred to a distinct soil package and a level was an arbitrary division within a stratum. Designed for more straightforward and unmodified archaeological sites, this system became untenable in the face of increasing complexity. In order to keep complicated site stratigraphy straight, a modified version of the Harris Matrix⁴ context system was employed.

Essentially, within each excavation area, every natural soil horizon, historical occupation surface, structure, feature (or part thereof), and stratigraphic division of fill sediment within a feature was assigned a unique serialized context number. These context numbers allow AECOM archaeologists to track, recreate in the laboratory, and analyze all parts of each site, down to individual fill layers. Although features were numbered sequentially in the field, some classes—most notably walls and postholes—were not consistently assigned numbers. During the process of compiling the geographic information systems (GIS) data from the field, the geospatial team therefore provided feature designations. The post-field feature assignments are sequential numbers preceded by the letters “GIS.”

Geomorphological Studies

Dr. Daniel Wagner of Geo-Sci Consultants and/or AECOM’s internal soils specialist, Elisabeth LaVigne, performed geomorphological examinations of intact soil profiles within the study area. The purpose of these studies was to characterize the development of preserved undisturbed soil horizons and to formulate interpretations regarding the age and nature of landform evolution within examined areas, to assess issues pertaining to landform stability over time, and to document the nature and extent of impacts to the local landscape during the historical era. Information generated through these studies assisted not only in the interpretation of any associated archaeological deposits, but also provided critical baseline data that future archaeologists can use to evaluate the cultural resource potential of this section of Philadelphia. The findings of all geomorphological investigations related to data recovery were fully documented in a separate report that is included as an appendix in associated Phase III reports of findings.

Site Mapping

All archaeological testing and identified cultural resources were intensively mapped using a combination of global positioning system (GPS) data collection with sub-meter accuracy, traditional hand-drawn plan view drawings, and total station data collection. GPS data mainly provided real-world reference points to tie in total station data to their real-world locations and to map discoveries made during archaeological monitoring efforts. For all non-monitoring efforts, a total station was used to establish excavation testing grids, identify individual test locations, and record the placement of archaeological resources within an established local site grid. Hand-drawn measured plans were generated during excavation as part of the data recordation and plans were drawn to include grid tie-in points, allowing them to be correlated with the site grid established using the total station. At the conclusion of fieldwork, these spatial data streams were combined into a singular GIS environment to provide a digital record of the real-world spatial placement of each archaeological survey area and record the exact location of each encountered archaeological feature. These digital spatial records of the excavation, in combination with other spatial data from georeferenced historical maps and aerials, were combined to conduct a variety of spatial analyses. The detailed mapping and digitization of the excavation proveniences (excavation units, shovel test pits, trenches, blocks, and features) enabled the study and visualization of artifact distributions, feature density, feature property association, and more.

Archaeological Laboratory Methods

Artifact Processing

Artifacts from the excavations, with the exception of faunal remains and some samples requiring outside professional expertise, were processed at either the AECOM archaeological laboratory in Burlington, New Jersey, or the I-95 Girard Avenue Interchange Archaeology Center in Philadelphia, Pennsylvania. Materials were washed, dry brushed, or separated for conservation as appropriate in accordance with the PHMC’s published collections and curation standards.⁵ Artifacts were separated by class and placed in individual 4-mil polyethylene bags labeled with provenience information using permanent marker. The individual bags were placed within a large bag(s) for an entire provenience. All bags were labeled and pierced for air circulation. These bags of artifacts, sorted by provenience, were then set aside to await cataloging and/or vesselization. Following cataloging, the collection was labeled in accordance with the PHMC’s published collections and curation standards.⁶ 

Cataloging and Computerization of Artifact Data

AECOM employs a Microsoft Access artifact inventory database to both collect and analyze archaeological collections. AECOM uses Access as the platform for its artifact inventory databases because this software provides the ability to tailor data collection to the project and its research goals while remaining compatible with other common collection management programs, such as the Re: Discovery context module or Past Perfect. AECOM maintains a daily computer backup file of all data on an off-site server to ensure data integrity.

After the initial washing and sorting processes, AECOM material specialists catalog the collection, directly inputting the attributes of each artifact into the Access database. This procedure reduces errors in data entry, as the individuals who analyze the artifacts also input the data. This eliminates the need to record on paper forms for later data entry by non-material specialists, saving time and eliminating a common source of transcription error from the cataloging process. To further ensure data integrity, AECOM’s artifact inventory database uses a data-entry form as the principle interface for data entry. The data-entry form can be used for both historical and pre-contact artifacts. This interface requires, at a minimum, information about artifact counts, temporal association (historic, prehistoric, both, or unknown), class, activity group (functional attributes following Hume⁷ and South⁸), material, object, and weight in grams. Weight is recorded because, for some artifacts (e.g., window glass; brick, mortar, and plaster fragments; and coal), this measurement can be more interpretively significant than simple count. For the sake of consistency and to ensure accurate queries, artifact attributes are drawn from a series of master lists via drop-down menus. These drop-down menus include standard terms (e.g., red-bodied earthenware, pearlware, etc.) to help classify different attributes of an artifact. While the system relies on master lists, it is flexible enough to add new artifact attributes and artifact types as the recovered material culture dictates.

In addition to the information required for all artifacts, other attributes are recorded, as appropriate. Additional attributes recorded for pre-contact artifacts include the raw material type, method of manufacture, shape, and use-wear. For historical artifacts, especially glass or ceramics, the database can record information about ware type, form, decoration, color, manufacturing technique, branding, and dates of production. Further information about part present, condition, wear, etc., can also be tabulated. Information about the relationship between artifacts via mending or crossmending is also recorded, and artifacts that all from part of a vessel are given a unique vessel number. This information can then be used to form the core of a minimum vessel inventory analysis for a given provenience. Additional fields are available for the description of glass and ceramic vessels, and the comments field is a memo field (i.e., unrestricted in length).

Once the artifact information is in the database, a variety of reports and queries can be generated using Microsoft Access, according to project needs. Microsoft Excel can then be used to manipulate the data and to generate tables, charts, and graphs, as needed. The artifact inventory database can be queried to conduct quantitative analyses to assess the attributes of the provenience from which material was recovered. Queried data is used to apply dating techniques like percent contribution, terminus post quem (TPQ), and mean ceramic date (MCD), as well as further define assemblages related to a provenience using minimum vessel inventories, minimum number of objects, and crossmending logs, which help ascertain the nature of deposits and their relationships to other deposits on the site. Such temporal and provenience data can then be combined with other database platforms like ArcGIS, where the spatial data from the excavation is linked to the quantitative artifact data to visualize artifact patterning and distributions across the site. Additionally, by linking in separate databases containing historical vital data from censuses, death records, deeds, and city directories, artifact data can be combined within the ArcGIS framework to examine artifact frequencies or distributions related to household or individuals of certain ethnic groups, occupations, religions, etc. This approach opens opportunities to create analytical queries that can help show temporal change or potentially highlight the material culture signature of groups.

Prioritization of Features

The extent to which artifacts were processed and analyzed was determined by the interpretive potential of the feature or deposit from which they were derived. The highest level of analysis was directed toward features and cultural deposits with high information potential, most notably shaft, pit, and midden features. Using the preliminary data gathered during both the excavations and the post-excavation artifact analyses, these high-information-potential feature types were divided into high-, medium-, and low-priority groups. For the purposes of lab processing, priority categories were not assigned to feature types such as foundations or posts, which typically do not have sufficient quantities of interpretable artifacts associated with them.

I-95 Feature Processing

Considering the large number of well-preserved archaeological features identified at many of the I-95 sites, intensive laboratory analysis is directed only toward features with exceptional information potential, namely shaft, pit, and midden features. These high-information-potential features are divided into high, medium, and low priority groups based on several criteria, and receive the analytical and processing treatment corresponding to each group.

Feature priorities should be set by first using field paperwork to determine intact deposits and primary contexts. An artifact bag review should also be completed (if no catalog yet exists) to identify what features contain diagnostic materials. High priority features that are somewhat redundant in assemblage may be sampled to only complete a full High Priority lab effort on a few chosen. Preliminary historic research may also be used to identify key properties and features that should be examined more fully.

High Priority Features

High priority features exhibit one or more undisturbed artifact-bearing deposits in which the original matrix, associations, and proveniences of the artifacts remain intact. These primary deposits can often be dated to a specific time period, related to a specific residence, and occasionally can be associated with particular families or individuals. Such connections allow archaeologists to make statements about class, economic status, and community standing (among other topics) with reference to specific households.

The lab’s mending and vesselization procedure is applied to all high priority features. Vessels are reconstructed as completely as possible in order to obtain minimum vessels counts (MNVs), vessel dimension measurements, and to document form and decoration as thoroughly and accurately as possible. The MNV for a feature or context refers to the minimum number of vessels that are represented by all the sherds in an artifact assemblage. MNV counts are a necessary analytical tool because it allows archaeologists to discuss the material in terms of vessels and objects (aka how people actually used them) rather than only sherd counts. The mending/vesselization of high priority features will be ended (and cataloging begun) as soon as no additional information can be drawn from additional mends (even if vessels aren’t entirely complete). This means that as soon as an accurate MNV count is attainable, vessel measurements possible, and form/decoration identifiable, mending for that particular assemblage/feature will stop. MNV counts are a necessary analytical tool because it allows archaeologists to discuss the material in terms of vessels and objects (aka how people actually used them) rather than only sherd counts.

The mending/vesselization of high priority features will be ended (and cataloging begun) as soon as no additional information can be drawn from additional mends (even if vessels aren’t entirely complete). This means that as soon as an accurate MNV count is attainable, vessel measurements possible, and form/decoration identifiable, mending for that particular assemblage/feature will stop.

Medium Priority Features

Medium priority features contain substantial numbers of chronologically diagnostic artifacts. However, the identification of mottled, inhomogeneous, or displaced soils; stratigraphic discontinuities; the admixture of temporally unrelated artefactual materials; and/or unexpected stratigraphic inversions within these features suggest that their contents were subjected to considerable post-depositional disturbance (e.g., multiple cleanout events, intentional filling with artifact-bearing soils, etc.).  Although such secondary deposits can provide useful information about the residents of the property or neighborhood in a general sense, they can less consistently be associated with specific residents or time periods.

The lab’s mending and vesselization procedure is also applied to all medium priority features, though the vesselization/mending of medium priority features is drastically shortened, likely limited to a day, at most, for the larger medium priority features.

Low Priority Features

Features classified as low priority were those that contained few or no chronologically diagnostic artifacts. At the end of their use-life as sanitary shafts, many privies were cleaned out and repurposed for the disposal of coal ash and furnace waste. Such features were bisected and, if no artifact-bearing deposits were encountered, excavation of the feature was terminated, leaving half of the feature unexcavated. In some cases, where such deposits were deep, only an initial half was typically excavated and then the remaining material was probed to determine if primary privy deposits or artifact-bearing secondary feature fills were present. If none were identified, the excavation was similarly terminated.

Low priority features contain few or no chronologically diagnostic artifacts. Shaft and pit features approaching the end of their use-life were sometimes cleaned out and repurposed for the disposal of sterile coal ash and furnace waste. Likewise, some refuse was deposited on and subsequently distributed across the ground surface producing extensive, albeit intermittent and inhomogeneous, sheets of sediment, ash, charcoal, and cinder mingled with incidental amounts of glass, ceramic, and bone. In either case these behaviors can create clearly identifiable archaeological deposits containing relatively little identifiable artefactual material.

Low priority features do not undergo any mending or vesselization, and will be analyzed in the lab on the same level as non-priority features (see below), with only raw counts (albeit still with accurate dating information) being entered into the catalog.

Non-Priority Features

Posts, trenches, walls, floors and other structural elements are also frequently encountered during excavations. While of limited analytical and interpretive value, these features serve to demarcate structural footprints, to illustrate the formal interconnectedness of various buildings that formerly occupied the site, and to establish when local residents adopted emergent construction, sanitary, and energy technologies.

Non-priority features are not mended or vesselized, but are merely entered into the cataloged as raw counts.

*Below is a breakdown of feature priorities for two large sites from the I-95 project. These numbers typify the project as a whole:

Gunnar’s Run South Site:

Total Features: 679

Prioritized Features: 105 (15.4% of Total)

                – High Priority Features: 27 (25.7% of Prioritized, 3.9% of Total)

                – Medium Priority Features: 22 (20.9% of Prioritized, 3.2% of Total)

                – Low Priority Features: 56 (53.3% of Prioritized, 8.2% of Total)

Fishtown 2 Site:

Total Features: 967

Prioritized Features: 180 (18.6% of Total)

                – High Priority Features: 46 (25.5% of Prioritized, 4.7% of Total)

                – Medium Priority Features: 59 (32.7% of Prioritized, 6.1% of Total)

                – Low Priority Features: 75 (41.6% of Prioritized, 7.7% of Total)

Both Sites Combined as Average/Sample of Larger I-95 Sites:

Total Features: 1,646

Prioritized Features: 285 (17.3% of Total)

                – High Priority Features: 73 (25.6% of Prioritized, 4.4% of Total)

                – Medium Priority Features: 81 (28.4% of Prioritized, 4.9% of Total)

*This means that 54% (@ half) of prioritized (High/Med/Low) features, and only 9% (@ 1 in 10) of total documented features will undergo some mending/vesselization

                – Low Priority Features: 131 (45.9% of Prioritized, 7.9% of Total)

Artifact Analyses

Pre-Contact Analyses

Lithic debitage (i.e., detritus from the manufacture of stone tools) was analyzed using a typological approach in order to better understand the types of lithic reduction activities occurring on site.⁹ Complete flakes and proximal flake fragments were assessed using characteristics such as flake morphology, striking platform type, termination type, dorsal flake scar count, and presence or absence of cortex. Based on these characteristics, debitage was categorized into the following technological types: decortication flakes, early reduction flakes, bifacial thinning flakes, bipolar flakes, and shatter.¹⁰ Flake fragments, which are defined as the distal, lateral, and medial sections of flakes, were considered non-diagnostic. Decortication flakes include specimens that exhibit nearly 100% cortex on the dorsal face and striking platform. Early-stage reduction flakes are characterized by relatively wide striking platforms that are often unprepared/flat or cortical, an exterior platform angle typically between 46º and 90º, pronounced bulbs of percussion, relatively flat longitudinal cross-sections, few dorsal scars (≤ 3), and the presence of cortex. Bifacial thinning flakes are defined based on the presence of striking platform faceting, grinding, and lipping, acute exterior platform angles (≤ 45º), diffuse bulbs of percussion, curved longitudinal cross-sections, several dorsal scars (≥ 3), feathered terminations, and little to no cortex. Bipolar flakes are lithic debris that may include all, or some, of the following characteristics: multiple fracture faces, wedge-shaped morphology, pronounced compression rings, and evidence of impact on opposite ends of the flake. Shatter is the term for blocky flake fragments—often associated with bipolar reduction—that do not have identifiable flake attributes, such as a platform, bulb of percussion, or ventral and dorsal surfaces.

Attributes recorded for debitage include the presence or absence of cortex, condition (i.e., whole or fragmentary), and evidence of thermal alteration. Cortex was classified as block, cobble, or absent. Blocky cortex consists of weathered rind and other coarse surfaces that are typically found on material recovered from primary outcrops. Cobble cortex is the rounded smooth surface found on natural cobbles. Thermal alteration of debitage was recorded as reddened, potlidded, crazed, or absent. Lithic raw material types were identified according to macroscopic characteristics such as color, translucency, and texture.

Hafted bifaces were classified using regional typologies outlined by Custer¹¹ (2001), Kinsey¹² (1972), and Ritchie¹³ (1971). Metric variables recorded include mass (grams), maximum thickness (millimeters), maximum length (millimeters), maximum width (millimeters), blade length (millimeters), shoulder width (millimeters), haft length (millimeters), neck width (millimeters), and basal width (millimeters).¹⁴8 Non-metric variables include raw material, thermal alteration, cortex type, cortex cover, haft shape, haft grinding, transverse cross-section, and blank type. These diagnostic features were used to determine the nature and ages of the various point types recovered from controlled excavations. This approach allowed for the establishment of a temporal range for the utilization of the area during the pre-contact period.

Pre-contact pottery was documented based on temper, surface treatment, rim forms, and (where possible) vessel size. All pottery was classified using standard regional typologies. As in the case of lithics, analyses of pottery vessels assisted in identifying activities that occurred at each site (e.g., cooking and storage). Examination of decorative motifs was used to help with intra- and inter-site analysis to determine if “related” social groups occupied sites in the same general area. Significant examples of pottery have been photographed or illustrated.

Historical-Era Analyses

At a minimum, basic analyses performed on recovered historical artifacts included the identification of key characteristics for each object, including general form and function (e.g., cut nail – architectural), material composition (ceramic, glass, metal, etc.), ware type (creamware, pearlware, whiteware, etc.), manufacturing technique, decoration, date of manufacture, and maker’s marks (if present), following accepted standards delineated in Hume,¹⁵ South,¹⁶ and Miller,¹⁷ among others. Dated artifacts were utilized to establish terminus post quem (TPQ), mean ceramic dates (MCDs), percent contributions, and minimal vessel counts for specified intact archaeological deposits. Artifacts recovered from intact and/or primary deposits additionally underwent intensive analyses designed to aid in the interpretation of these artifacts and the contexts in which they were found, as well as to help answer the research questions developed for this project. In particular, glass and ceramic vessels, as well as select small finds (smoking pipes, toothbrushes, etc.), were crossmended both within and between appropriate depositional contexts. Additional artifact characteristics were recorded for identified vessels, including those related to details of decoration, use-wear, specific functions, and, for bottles, any information about their contents. Comparative statistical data were generated for artifacts recovered from each identified depositional/study unit within discrete features.

Terminus Post Quem

Terminus post quem (TPQ), which translates to the “date after which,” is a dating technique used to determine the earliest possible date for a given provenience based on the most recently produced artifact within it. If the deposit has not been subject to subsequent disturbances, it can only have been created after that object was first produced.

Mean Ceramic Dating

Mean ceramic dating is an analytical technique Stanley South pioneered.¹⁸ This technique is accomplished by using the following formula:

MCD=\frac{\sum_{i=1}^{n}{X{i}}{f{i}}}{\sum_{i=1}^{n}{f{i}}}

In this formula, X_i equals the median manufacturing date of a particular ceramic type and f_i equals the frequency (count) of that ceramic type within a collection.

\sum_{i=1}^{n}{X{i}}{f{i}}

This portion of the equation represents the sum of all the median manufacturing dates multiplied by their respective frequency.

For instance, if you have 3 sherds of earthenware with a median date of 1780 and 2 sherds of stoneware with a date of 1795, the calculation would be: ((3*1780) +(2*1795)).

\sum_{i=1}^{n}{f{i}}

This portion of the equation is the total number of sherds being analyzed. In the above example, this value would be (2+3), giving a value of 5.

Continuing this example, the final calculation would be ((3*1780) +(2*1795))/5, yielding an MCD of 1786.

While this technique is useful for establishing a midrange for an entire deposit, it does not guarantee that a deposit specifically dates to that time. In the previous example, the MCD is 1786, but based on manufacturer, design, or decoration, the refined earthenware might have a TPQ of 1790, which would mean that the MCD is predicting a date earlier than the earliest manufacturing date for an item in the deposit. Thus, the MCD technique cannot be used to the exclusion of the TPQ. This technique can, however, compensate for potential outliers in the TPQ that might otherwise skew the dating of a context.

Due to the nature of manufacturing dates from the mid-nineteenth through early twentieth centuries, both beginning and ending mean dates were calculated for contexts, thus generating a mean (or theoretical peak production) range rather than a single mean date. This approach allowed for the inclusion of significant beginning OR ending dates for specific artifacts in the mean range calculation, while omitting beginning or ending dates that would otherwise be statistical outliers, frequently either an ending date extending to present or a beginning date far preceding the approximate known dates of the context. Each context mean range was then used in conjunction with a percent contribution created from the same data set and the context TPQ in any ultimate interpretation.  

Percent Contribution

Percent contribution is a refinement of Stanley South’s ¹⁹ original mean ceramic date calculations. It is useful in understanding occupation peaks across sites because it shows a range rather than a single date. The percent contribution indicates the probability of a randomly selected sherd from a particular provenience being manufactured in a given year. The method used to create this chart is found in A. F. Bartovics’ The Archaeology of Daniels Village: An Experiment in Settlement Archaeology.²⁰ The formula used is:

P = S/(N*D)

Where:

P is determined for each ware type with a unique date range (for example, 1744–1775 for scratch-blue decorated white salt-glazed stoneware). The value is then entered into each year of manufacture for that ware type. Each year’s cumulative probability is determined by adding all the values of P for each ware type manufactured in that year. This cumulative percent value is then graphed for the range of years.

For example, a 10-sherd collection dataset (as shown in Table 1) would yield the percent contribution chart Figure 1, which represents the likelihood that any artifact in the collection was deposited in a given year. While the overall date range of the assemblage spans the years 1794–1840, the peak probability occurs between 1820 and 1830, suggesting that this is the most likely depositional date range for the assemblage.

Percent contributions and mean ranges are mutually informative in the dating of a context. While a single mean range obviously assumes a single peak in the distribution of the data (a single or more than one otherwise tightly dated depositional events), a percent contribution graph will depict the true distribution of the dates and show multiple temporally distinct deposits, if present. By overlaying an assemblage’s mean range over its percent contribution, the probability that a randomly selected artifact was made during the mean range can be calculated by summing the probabilities of all the years within the mean range. A relatively low probability will obviously convey that the mean range is not adequately representative of the assemblage as a whole, with the mean ranges typically representing valleys between peaks in such instances (in other words, very little of the actual assemblage at all). In cases where the mean range does not sufficiently represent the actual data, a context’s date range (or more likely, multiple peak ranges) is derived from the graph itself. The interpretation of a percent contribution graph is obviously subjective, and thus the dates pulled from such graphs can vary slightly from person to person. As such, where a mean date range (mean beginning – mean ending) for a context adequately represents a percent contribution distribution with a single peak, the more objective, calculated mean dates take precedence and are used as the context dates.

Minimum Vessel Inventory

A minimum vessel inventory analysis is used to establish the minimum number of vessels (MNV) represented within an assemblage. This technique diverges from the aforementioned strategies, as they rely on sherd frequency, not vessels. Knowing the number of vessels represented in a collection can help account for sample bias imposed by sherd counts. Variability of breakage or recovery can artificially inflate the importance of a ceramic group. Vesselizing via crossmending helps to control for this bias. For example, a redware plate that broke into two sherds will statistically appear less significant in an assemblage versus a whiteware plate that broke into 30 sherds. By vesselizing, it is possible to determine that only one vessel of each ware type is present, thus putting redware and whiteware on equal footing in the assemblage. Vesselization also allows for a more accurate determination of vessel form, which provides a better sense of activity areas and foodways. Additionally, vesselization can help determine the kind of deposit present. A high MNV count with a low percentage of complete vessels can indicate a deposit that has faced significant post-depositional movement, while a low number of vessels that are mostly complete can indicate an assemblage that remained largely untouched after deposition. Finally, vesselization aids in the stratigraphic understanding of a feature. Although contexts within a feature may appear physically different, crossmends between them can indicate that they were deposited at or around the same time. This allows contexts within a feature to be grouped into analytical units, thereby facilitating feature interpretation.

The minimum vessel counts were done by feature and took place after mending. There are two predominant methods for conducting MNV counts: quantitative and qualitative. Quantitative MNV counts are based exclusively on rims, bases, and handles, largely discounting body sherds. The advantage of the quantitative method is that it is highly replicable; however, it may result in a disproportionately low vessel count because body sherds are not considered. Qualitative MNV counts are more subjective and therefore less replicable, but they allow analysts to consider multiple attributes resulting in a more accurate count.²¹ The AECOM lab staff employ the qualitative method. To mitigate the subjectivity of the qualitative method, lab staff consistently adhered to the following guidelines.

When possible, a single analyst performed MNV counts. When multiple people worked on the same MNV count, all analysts used the same parameters and agreed on all MNVs. For each sherd, there were three possibilities: a sherd could be assigned to a group of sherds representing an existing MNV with shared attributes, a sherd had unique attributes that indicated it should get an MNV, or a sherd had attributes that would allow it to be assigned to one or more groupings. These sherds were not considered in the MNV count. For ceramics, all sherds within a ware type were considered and further narrowed down based on decoration and object form.

MNV counts for glass vessels were trickier because glass often lacks the decoration that is crucial in ceramic MNV counts. Generally, nondescript glass body sherds were discounted unless they represented a unique sherd in the assemblage. For example, the only sherd of amber colored glass in an assemblage would get an MNV. Finishes/rims and bases were considered when conducting glass MNV counts. Complete finishes/rims or bases were given an MNV unless there were multiple complete finishes/rims or bases that were potentially part of the same vessel, which could result in a vessel being counted twice. Lab staff worked through this issue by laying out the assemblage and determining if there were more complete finishes/rims or bases and then discounting whichever sherd type amounted to fewer pieces. For example, if there were four aqua beer bottle finishes but six aqua beer bottle bases, each base would be given an MNV and the finishes would be discounted. If the analyst could be certain that finishes/rims and bases did not belong to the same vessel, both were given an MNV. Such interpretations were subjective and varied by glass color and vessel type.

During cataloging, each vessel was given a “percent complete” designation, which aids in determining the kind of deposit present (e.g., primary versus secondary). The Microsoft Access database used in the cataloging of this project has nine categories: 0–2%, 3–10%, 11–25%, 26–50%, 51–75%, 76–95%, 96–99%, 100% complete, and 100% mendable. For this project, categories 0–2% and 3–10% were combined for analytical purposes.

Specialized Studies

Faunal Analysis

Faunal remains (archaeologically recovered animal bones) have been collected and analyzed throughout the project to address a wide range of research topics, including foodways of the inhabitants of Philadelphia, food landscape, food chain, the existence of commensal and pet species in the urban community, etc.

Methodology

Account of the Comparative Collections

Comparative zooarchaeological collections utilized for the faunal analyses described here include those housed in the Zooarchaeology Laboratory in the University of Pennsylvania Museum of Archaeology and Anthropology (Penn Museum) and materials from the Academy of Natural Sciences in Philadelphia (ANSP), as well as numerous published references.²² A team of zooarchaeologists analyzed the faunal materials: Teagan Schweitzer, Ph.D., who studied the collection at the Penn Museum and ANSP, and Marie-Lorraine Pipes, Ph.D., at SUNY Geneseo, New York. The faunal inventory was entered into Microsoft Access and integrated into the main artifact database.

Identification

The assemblages discussed here were analyzed using conventional zooarchaeological techniques.²³ Whenever possible, each bone or bone fragment was identified to species level. When this level of identification was not possible, a series of higher-order taxonomic categories was used. The sheep/goat category was used for almost all caprine remains. Historical information indicates that sheep were much more common in Pennsylvania in the eighteenth and nineteenth centuries. Additionally, no bones were identified as having specifically goat-like characteristics.²⁴ Thus, it is assumed that most of the bones in the sheep/goat category belonged to sheep. When bones could not be identified to either species or family, they were assigned to categories such as large, medium, or small mammal, explained further below. Other small bone fragments were simply assigned to class (e.g., mammal, bird, or fish). A very small number of bone fragments remained entirely unidentifiable.

Mammal and bird bones in these assemblages were placed into designated size classes. This technique facilitated the classification of bones that could not be identified to a more specific taxonomic level (see Table 2 below). For specimens that could not definitively be assigned one size or another, both potential sizes are listed.

Analysts recorded faunal catalogs in Microsoft Excel documents and later imported them into a Microsoft Access database to be integrated with other material and provenience catalogs. Aside from the identification of each bone or bone fragment to the most specific taxonomic degree possible, bones were further identified by element, side, completeness (fragment or whole), state of fusion (fused or unfused), weight (in grams), cut marks, gnaw marks, degree of weathering, heat exposure (burned black or calcined), relevant measurements,²⁵ and aging based on dental eruption and wear.²⁶ Measurements were taken using a dial caliper to the tenth of a millimeter. Most of the measurements in the faunal assemblages were taken on bird bones because they are more often recovered in a complete form in the historical period archaeological record. Several measurements were also taken on sheep/goat long bones due to their not insignificant representation among the faunal remains. Cattle and pig bone measurements were most often applicable only to the bones of the lower limbs (i.e., metapodials, carpals, tarsals, and phalanges), as these elements are smaller than many of the others and more often remain intact even after butchering.

Quantification Techniques

Units of quantification used in these analyses include the number of identified specimens (NISP), the minimum number of individuals (MNI), and specimen weight (in grams). The techniques involved in these quantification units are subject to critique, but, when taken together, they facilitate a more comprehensive assessment of the total faunal assemblage.

The NISP is a count of all identifiable bone fragments. It is used to assess the relative abundance of a particular species within an assemblage. The main problem and the implicit assumption embedded within NISP calculations is the presupposition that each bone or bone fragment originated from a separate individual. However, some—if not many—of the bones in the assemblages may have derived from the same animal or same cut of meat and would consequently be counted numerous times when using this quantification technique. This issue is referred to as the problem of interdependence. Additional concerns regarding this technique are that it does not consider differential preservation, differences in collection techniques, or variability in the identifiability of certain skeletal elements.²⁷

MNI calculations, which estimate the smallest number of individual specimens that would be necessary to account for all the bones represented in a given assemblage, often counterbalance NISP calculations in faunal analysis. This technique estimates the relative importance of the various species found present by using the most commonly identified element in each species. It provides a conservative estimate of the number of animals represented by the bones in any given assemblage. The problem with MNI is that it can result in an overestimation of the importance of the less common species in an assemblage. Small sample sizes, in particular, tend to overemphasize the relative abundance of the rare or less abundant species within a particular sample.²⁸

Problems with NISP and MNI can be exacerbated in historical faunal assemblages because smaller fauna such as fish or birds were often sold whole, whereas larger animals such as cattle, pigs, and sheep were often, but not always, sold as individual cuts of meat. Consequently, preference for a particular cut of meat might result in overrepresentation of the MNI of a particular species in an assemblage.²⁹

Another method of quantification is the calculation of weight. As Terry O’Connor describes, this technique rests on the notions that “a predictable and sufficiently constant proportion of the weight of a mammal is its skeleton, and another predictable and sufficiently constant proportion is potentially edible muscle.”³⁰ Given these assumptions, comparing the relative abundance of taxa by weight of bone fragments should provide an estimate of their potential dietary contribution. Nonetheless, many varied concerns have been raised over the use of weight as a valuable quantification technique. For example, differential destruction and transport of bone fragments of different densities can skew the composition of an assemblage. Additionally, the bone weight to total body weight ratio is not consistent throughout the animal kingdom.³¹ However, comparing the weights of the bones of animals in the same class with relatively similar body sizes should allow the analyst to estimate potential dietary contribution.

Ultimately, it is the combination of all these techniques that yields the clearest picture of the meaning of faunal remains in any given assemblage. Incorporating multiple methods while keeping in mind each of their strengths and weaknesses culminates in the best possible understanding of the zooarchaeological remains.

Flotation-Recovered Faunal Material

Animals recovered from flotation can be useful to the archaeologist because they allow for the collection and identification of smaller and lighter bones than are typically found in material screened through 0.25-inch mesh. The hope is that by analyzing floats, a more complete picture can be recreated of the faunal landscape and the uses of various animal species. Some species are much more likely to appear in and be identified in flotation samples, such as small or immature fish, small mammals (e.g., mice, rats), and small amphibians or reptiles.³² Many more fish scales, for example, are recovered from flotation than from standard screened material in the field. Without the flotation data, therefore, the archaeologist is left with a potentially incomplete view of the animals that were part of the overall system being represented among the archaeological remains.

Rather surprisingly, not many other projects explicitly incorporate much of the flotation material into their faunal analysis. This could be, in part, because analyzing these materials is extremely labor intensive. Projects also do not always have a standard strategy for recovering flotation samples and therefore their comparative usefulness is potentially limited, as in the case of the I-95 project.

Rebecca Yamin, in her report on the faunal methods used for analyzing materials from the Five Points neighborhood in New York, mentions that “the spines and scales of unidentified fish are excluded from tabulations and calculations contained in this section so as not to inflate the numbers of unidentified fish in these deposits.”³³ She notes that morphological similarities between flounder and bass species made higher (to species) level identification difficult for many of the bones. There were also quite a large number of flounder and bass species available in the local landscape, adding another layer of difficulty to making species-level identification. Most of the time, these bones were identified as larger bass or flat fish.³⁴ What Yamin does not say is whether any of the fish materials came from flotation samples.

Claudia Milne mentions in her report on Area F, Site 36PH0011, of the Independence National Historical Park  Excavation in Philadelphia that herring bones were likely missing from the assemblage because of lack of fine mesh or water screening.³⁵ This difference relative to Yamin’s report is notable, but it is likely that the lack of flotation samples included in the analysis missed additional important fish-related and other species information as well. The results from later work in Philadelphia at the Museum of the American Revolution Site also make no mention as to the incorporation of material from flotation in the analysis.³⁶

As David B. Landon (2005) states, flotation samples can be helpful in establishing a greater understanding of the taphonomic effects of the depositional environment, which, in turn, can have a significant impact on the faunal remains (presence/absence, size of remains, etc.).³⁷ He notes that Robert L. Jolley made the point that “sample size, recovery methods, preservation factors, and modification of the faunal assemblage by natural and cultural factors” are rarely considered in studies of historical assemblages and encourages historical archaeologists to make this more of a priority.³⁸

In general, very few specifics are provided in the methods sections of comparable reports about flotation use or collection and processing. It seems that this recovery technique has still not reached its potential in terms of providing useful data to zooarchaeologists.

Faunal Analysis of Flotation Materials

Similar to the rest of the faunal materials identified during the project, the specimens recovered from floated soil samples were identified to the closest taxonomic level possible using comparative materials and zooarchaeological manuals (see above). Many, but not all, of the faunal remains from the flotation samples were analyzed. Some have been reserved for future analysis as needed or desired. In general, because of the size of the remains recovered in flotation samples, many of the individual bone fragments are not identifiable to species level. Thus, they were categorized into unidentified large, medium, or small; mammal, bird, reptile, fish, or amphibian. Bones lumped into these larger categories were weighed to provide a rough estimate of quantity but were otherwise left uncounted. AECOM archaeologists determined that the additional labor required to count these undiagnostic specimens would not produce additional meaningful data for most analyses. Where more specific identification was possible, counts were included in the catalog. However, the individual elements were not cataloged separately. Instead, an overall count per species was cataloged with some general comments about the elements present in the notes field. These abbreviated cataloging methodologies allowed AECOM archaeologists to efficiently summarize this large portion of the assemblage, which historically has been ignored.

Not all fauna-bearing contexts in the I-95 project area resulted in the collection of flotation samples. Because of this discrepancy in recovery methodologies, putting the flotation material directly into the discussion and comparison of faunal assemblages could be problematic. Different recovery strategies are likely to result in differently characterized collections of material, and it would be a mistake to conflate variation in modern archaeology with variation in historical dietary and depositional activities. In light of this, AECOM zooarchaeolologists attempted to be explicit as to where and when flotation material was included in the description of an assemblage. Unlike many other archaeological projects, work for the I-95 project included cataloging, analyzing, and incorporating flotation samples whenever possible and relevant to the data interpretation and discussions. If the dataset had been ignored, an incomplete understanding of historical foodways might result. By including such data, archaeologists are able to identify and discuss portions of the diet that might otherwise have been missed.

Archaeobotanical Analyses

Archaeobotanicals (archaeologically recovered plant material) have been collected and analyzed throughout the project to address a wide range of research questions, including health and dietary characteristics, community and ethnic identity, resource access and trade networks, and reconstruction of the local environment. The vast majority of archaeobotanical data from the project consists of macrobotanicals collected from soil samples.

Soil samples collected during the project were analyzed for macrobotanical remains (e.g., seeds, stems, skins visible with a low-powered microscope). Unlike microbotanical remains such as pollen, macrobotanicals do not generally travel far from their original source unless transported, either by humans or animals. Therefore, any macrobotanical remains recovered from an archaeological context are assumed to either result from human or animal activity or growing in close proximity to the area in which they were eventually recovered. Once deposited, the depositional environment of a site plays a large role in the survival of macrobotanical material. Botanical material tends to survive archaeologically in three ways: by being charred, being deposited in an anaerobic environment, or by being deposited in an arid environment.³⁹ The historical botanical material recovered during the I-95 Girard Avenue Interchange Improvement Project typically survived in anaerobic privy environments, whereas botanical material recovered from pre-contact contexts tended to be carbonized.

Soil Sampling and Flotation

During excavation, soil samples were primarily recovered from culturally significant features or features with a final depth of at least 3 feet (1 meter). A standard quantity of at least 2 liters of unscreened soil was collected in the field and brought back to the laboratory for flotation and analysis. Two methods were utilized to float soil samples during the project: barrel flotation and bucket flotation. Barrel flotation was utilized to process the vast majority of samples, while bucket flotation was used to process samples that were small or contained delicate plant material.

Macrobotanical samples were floated using a modification of the procedures Sarah Peterson outlined.⁴⁰ Soils recovered during excavation were weighed and then separated into light and heavy fractions using a modified flotation process. The sediment that resulted from the flotation process was discarded. The respective fractions were air dried. A minimum of 0.5 liters of soil from potentially rich or significant contexts was retained for possible future analysis. Non-archaeobotanical artifacts recovered from soil samples were prepared for subsequent identification and analyses.

Macrobotanical Analysis

Once dry, the light and heavy fractions were passed through nested geological sieves (3.35 millimeters, 1.7 millimeters, 1 millimeter, and bottom pan), creating manageable analytical separations referred to as splits. Each series of splits was weighed and then analyzed using a low-powered light stereomicroscope (6.3–40x). Seeds, charcoal, and plant parts were removed and identified to the closest taxonomic level using AECOM Burlington’s botanical reference collection, as well as manuals for the identification of seeds and plant parts.⁴¹ Identified plant specimens were recorded and weighed. Once identified, macrobotanicals were cataloged into site-specific Microsoft Access databases with the rest of the general artifact assemblage. Macrobotanicals collected during artifact screening were similarly identified to the closest taxonomic level and cataloged in the site-specific Microsoft Access database. AECOM archaeobotanical specialists analyzed the majority of soil samples for macrobotanicals, with select samples sent to Justine W. McKnight, archaeobotanical consultant.

Charcoal Analysis

When applicable, a charcoal sample analysis was conducted on pre-contact features. The methodology for charcoal analysis consisted of collecting 25–30 charcoal specimens larger than 1.7 millimeters and identifying each specimen to the closest taxonomic level. Each sample was counted, and then total weights were taken for each identified taxon. In the case of samples that did not contain 25–30 charcoal specimens, as many fragments were identified as possible.

Wood Identification

Several architectural wood specimens underwent taxonomic identification. Identifications were made at the AECOM Burlington archaeological laboratory and by Dr. Heather Trigg at the University of Massachusetts, Boston.

Microbotanical Analysis

PaleoResearch Institute in Golden, Colorado, conducted specialized analyses on column soil samples removed from intact cultural and non-cultural horizons and pre-contact hearths to identify pollen and phytolith remains. PaleoResearch similarly utilized organic residue analysis on sizable sherds of pre-contact pottery or projectile points to isolate and identify pollen, phytoliths, and starch grains.

Interpreting Archaeobotanicals

Analytical findings were utilized to address issues relating to diet, health, food preference, cultural and community identity, environmental reconstruction, and trade networks. Archaeological reports, ethnographic data, and contemporary literature—such as historical cookbooks, gardening catalogs, maps, photographs, and newspapers—were actively used to contextualize identified archaeobotanicals within the surrounding landscape. Interpretations of the archaeobotanicals recovered during the project came with several considerations. Macrobotanical material recovered from privies often consists of ambient seed rain, seeds eaten and deposited as fecal matter or “nightsoil,” and trash deposited from yard cleanings and kitchen refuse. Carbonized macrobotanicals had to have been introduced into an environment where they were burned but not incinerated prior to deposition. Seeds and plant parts deposited from these activities only represent a small portion of plants that may have been present in the surrounding environment or used by humans for food, medicine, and decoration. Some plants used for food are harvested prior to producing seeds or are processed in a way that destroys and/or removes seeds. The recovery of microbotanicals such as pollen, phytoliths, and starch grains can be similarly hindered due to a variety of reasons, including methods of pollen dispersal, cultural plant uses, and site-formation processes. In addition to depositional biases, the nature of certain post-depositional environments results in not all botanical material being preserved archaeologically. This means that a significant percentage of plants used by the local community or present in the surrounding environment, if deposited archaeologically, would not have been recovered. Interpretations of archaeobotanical material recovered during the project factor in the potential for missing data.

Radiocarbon Dating

Carbonized wood samples recovered from pre-contact feature excavations were submitted to Beta-Analytic, Inc., for radiocarbon dating. One sample from each feature level containing such material was selected for analysis.

Residue Analyses of Pre-Contact Artifacts

Any sizeable sherds of pre-contact pottery or projectile points that exhibited evidence of remnant materials on a surface was sent to PaleoResearch Institute for analysis. Their team used advanced laboratory procedures to remove and analyze any pollen, phytolith, starch, or protein residue present.

Collections Management

The scale of this project (over 2,000,000 artifacts as of December 2019) has necessitated the creation of a data-driven collections management system to track locations and statuses of all parts of the assemblage. All artifact boxes are barcoded and then recorded in a dynamic mobile app, which AECOM has designed in-house to be straightforward but also meet specific needs—offering the ability to track parts of the collection as they are washed, cataloged, mended, or removed for specialized analysis. This system is set up so that the whole lab process can be tracked by site, feature, or context-level FS number. This data can be used in concert with artifact-level catalog data for the larger-scale analysis of project data.

Photo-Documentation of the Artifact Assemblage

Selected diagnostic artifacts were photo-documented during the analytic process. Photographs were taken primarily for two reasons: 1) to use in reports, exhibitions, brochures, professional presentations, and other public-outreach projects and 2) to provide a visual record of some of the artifacts. The photographs consist primarily of high-resolution digital color images, with some black-and-white prints taken as needed for recordation purposes. The artifacts chosen for photography are from significant contexts or represent noteworthy examples of particular classes of material, styles, or manufacturing techniques.

Artifact Conservation

The laboratory staff assessed the physical, chemical, and biological conditions of the artifacts. Passive conservation measures, such as the use of proper storage bags and archival materials, were applied to the entire collection. A select group of diagnostic artifacts were sent to the Jefferson Patterson Park & Museum conservation laboratory for cleaning, stabilization, and preservation.

Artifact Curation

All artifacts were packed for permanent curation following the PHMC curation guidelines.⁴² Artifacts were packed using only acid-free, durable materials. Unless other repositories are identified, the Pennsylvania State Museum will receive all project-related products, including the artifact collection, final report, and additional project documents, including field notes, field records, maps, and photographs.

Select collections or assemblages of recovered artifacts may be set aside and separately curated for eventual long-term display/exhibit at one or more cultural repositories in Philadelphia. The development of any such select exhibits will be coordinated with project officials and the staff of the Pennsylvania State Museum and prepared in accordance with current museum guidelines.

I-95 Artifact Processing

The following describes the step-by-step process undertaken by the lab after dirty artifacts arrive from the I-95 field efforts to either the Burlington, New Jersey or Philadelphia lab facilities. Much of this text has been adapted from the actual training manual currently used by AECOM lab staff.

*NOTE:

• There is currently no formal artifact culling agreement in place with the SHPO as part of the I-95 project.
• However, the field crew undertaking the archaeological excavations along the I-95 corridor is composed of trained professionals who have the ability/knowledge (and have been doing so for some time) to take much smaller, representative samples of more insignificant artifacts that are typically recovered in very large quantities and provide little interpretive value.
• Such artifacts include clam/oyster shell, window glass, coal, and coal ash/cinders, among others.
• That being said, ALL artifacts that ultimately make it to the lab are then fully processed. To this point no culling of artifacts has taken place in the lab itself.

Washing Artifacts

• Before washing a bag of artifacts begins, the information on the white provenience tag (filled out in the field) inside the bag of artifacts is checked against the provenience information written on the bag itself for any inconsistencies. Only after this check is completed can the original, dirty field bag be discarded.
• The FS number, or Field Sample Number, is found on both the tag and bag. This is the unique number given to each separate provenience in the field. This number is linked to specific spatial data, both horizontally across the site and vertically within the ground/archaeological feature.
• Artifacts are washed (when appropriate) using a soft-bristled toothbrush with water and small amounts of mild soap.
• Dirty artifacts are never dumped directly from the field bags into water, as some artifacts should not soak in water due to their material or decoration type. Artifacts are always placed by hand into the washing tubs and carefully inspected while doing so.
• For artifacts with difficult to reach places like complete bottles and pipe stems, special tools such as bottle brushes and an ultrasonic machine are used to remove all dirt.

Drying Artifacts

• Washed artifacts are placed on paper towel-lined, perforated trays to dry. These trays are stored on portable racks, and each tray is labeled with the project name and I-95 project section.
• The provenience tag that came in with the artifacts from the field is kept with the artifacts on the tray while the artifacts are drying.

Re-Bagging and Re-Boxing Cleaned Artifacts

• A tray of cleaned artifacts is only re-bagged after completely dry.
• When re-bagged, artifacts are separated by material (glass, ceramic, metal, shell, etc.). This saves time when artifacts are brought back out to process and catalog by material type.
• Each new bag contains the same provenience tag (or new copy if the field tag needs replacement) as was with the dirty artifacts and on the drying tray.
• Each new bag has the following information written on the exterior:
  • I-95/Girard
  • Project Section #
  • Site Name
  • FS #
• Faunal material (animal bone) is pulled from its original drying tray and bagged separately for analysis by AECOM specialists (see below).
• When faunal material is separated from its original bag/tray, a “pull tag” is inserted into the original bag listed what was removed, the quantity, and by whom.
• Cleaned and bagged artifacts are placed into Bankers or Staples boxes, not archival level Hollinger boxes. Hollinger boxes are for artifacts that have already been cataloged and are being rehoused for final curation.
• Cleaned artifact bags are lined up in the boxes, not stacked on top of each other. Though this means fitting fewer bags in each box, it is safer for the artifacts as they are typically stored in these boxes for a lengthy period of time before being further processed.
• Cleaned artifact boxes are labeled with the following information:
  • I-95/Girard
  • Project Section #
  • Site Name
  • Feature #/Excavation Unit #
• Cleaned faunal material is boxed together and given a similar label as the regular artifact boxes.

*Artifacts for Conservation

• Artifacts considered potentially worthy of conservation are not washed and re-bagged/re-boxed in the manner of other artifacts.
  • If a potentially conservable artifact arrives at the lab wet, it is placed in a clean bag, sealed, and kept wet.
  • If a potentially conservable artifact arrives at the lab dry, it is not washed, but rather kept dry and sealed in a clean bag.
  • These artifacts are kept sealed in the state in which they were brought in from the field until further decisions are made regarding their conservation eligibility.
    • Artifacts deemed eligible are conserved by the Maryland Archaeological Conservation (MAC) Laboratory.
• Examples of common conservable artifacts for each material type:
  • Wood: tool handles, pulley parts, furniture parts
  • Softer Rubber: toy balls, wheels, hose/tubing, mirrors (this does not include hard rubber combs, buttons, etc.)
  • Leather: shoes (and shoe parts), belts
  • Metal: hardware, buckles, buttons, tool heads
  • Tortoise Shell: hair combs/personal items
  • Modified bone: bone buttons, utensil handles, syringes, umbrella parts
  • Modified shell: shell buttons
  • Cloth: clothing items such as shirts, socks, hats

Feature Mending & Visualization

• The AECOM archaeology lab mends (re-fits) and vesselizes ceramics and glass from all medium and high priority features (typically privy shafts) from the I-95/Girard project.
  • Artifact mending makes up the bulk of the feature mending and vesselization.
  • Due to layout space constraints, glass and ceramics artifacts from the same feature are typically mended separately.
• To begin mending, layout tables are divided by FS # (Field Sample #), using masking tape to clearly mark separate spaces for each FS #.
  • All vessel ceramic/glass artifacts are removed from their clean provenience bags and placed within the space on the table that corresponds to their FS #.
  • Once all of the ceramic/glass artifacts have been removed from their bags, each sherd is hand labeled with their FS # using masking tape and a fine-tipped sharpie.
  • After all of the sherds have been labeled with their FS #, they can be freely moved around the layout tables to begin mending.
  • Mends are secured using masking tape (never scotch tape) during this process.
• When mending has been completed for a feature or context, Minimum Vessel Counts (aka Minimum Number or Vessels or MNVs) are assigned to the assemblage.
  • MNV counts refer to the minimum number of vessels (and the minimum number of each vessel form) that are represented by the sherds in an archaeological assemblage.
• Mending and Vesselization (assigning of MNVs) is needed to allow archaeologists to interpret and discuss archaeological deposits on the object level rather than the sherd level.
  • Dating a deposit based on MNVs rather than sherd counts is more accurate because while one vessel may have been recovered complete, another may have broken into many pieces, which in turn would skew the statistical dating process.
  • Mending vessels allows archaeologists to better observe and record vessel shapes, dimensions and decorations, and in turn use those attributes in dating and interpreting the vessels.
  • Mending vessels allows archaeologists to examine and understand the relationships between different strata/levels/contexts with the same feature and in turn interpret that feature.
    • The presence of vessels that mend across different levels of the same feature could indicate a close temporal or other contextual link between those levels (and the opposite in the lack of mends between two levels).
  • Most importantly, people in the past used vessels and objects, not sherds. Assigning MNVs to archaeological assemblages allows archaeologists to speak anthropologically about how people lived in the past.

Artifact Cataloging

• All of the artifacts from the I-95/Girard project are cataloged in a Microsoft Access database.
• All FS logs (logs of the FS #s and their corresponding provenience information) are entered into the Access database for the project and are consulted if provenience questions arise.
• Each artifact that is cataloged is assigned an arbitrary entry number that is written on its bag and used as a unique identifier for that artifact.
  • An example of an I-95/Girard Project catalog #:
    • Example: “6.3000.1”
      • The “6” refers to the project Section #.
      • The “3000” is the FS #.
      • The “1” is the Entry #.

Artifact Marking for Final Curation

• Following cataloging, all artifacts larger than a square inch in size from the I-95/Girard Project are marked with their Site # and Catalog # for final curation and according to PHMC standards.
• Base coats and top coats consist of a mixture of B-72 pellets and acetone.
  • All artifact marking is done on a base coat, then sealed with a top coat once dry.
• Artifacts are marked with black or white India ink using calligraphy pens.
  • Black ink is used to label artifacts with light surfaces.
  • White ink is used to label artifacts with dark surfaces.
• The following types of artifacts are not marked:
  • Fabric
  • Leather
  • Mortar
  • Metal
  • Seeds & Nuts
  • Shell
  • Unfinished Wood
• Any bags containing artifacts smaller than a square inch, or consisting of a material that does not get marked, get a hand-written tag, with the “Site #/Catalog #” on acid free paper.
  • Calligraphy ink is used on hand-written tags.

Final Curation & Collections Management

• Artifacts are housed in acid-free Hollinger Metal Edge boxes for final curation with PHMC.
• AECOM has implemented a barcoding system to manage the 3,076 (and counting) Hollinger boxes and oversized artifacts that currently constitute the I-95 collection.
  • This allows for the tracking of box inventory, processing status, and physical location of the entirety of the collection at any given time during the project’s ongoing efforts.

*Specialist Studies – Archaeobotanical Analyses

• Archaeobotanicals (archaeologically recovered plant material) are collected and analyzed from the I-95/Girard Project to address a wide variety of research questions, including health and dietary characteristics, community and ethnic identity, resource access and trade networks, and reconstruction of local environment.
• The vast majority of archaeobotanical data analyzed by AECOM Archaeobotanical Specialists consist of macrobotanicals collected from soil samples.
  • Soil samples are analyzed for macrobotanical remains (seeds, stems, skins, etc. visible with a low-powered microscope).
• During I-95/Girard excavations, soil samples are typically recovered from culturally significant features such as privy shafts, pits, and hearth features.
  • A standard quantity of at least 2.0 liters of un-screened soil is collected in the field and brought back to the laboratory for flotation and analysis.
  • The flotation process separates the macrobotanical samples into light and heavy fractions, which, after dried, are examined for identification and analyses.
• Once identified to the closest taxonomic level, macrobotanicals are cataloged into the I-95/Girard Microsoft Access database with the rest of the general artifact assemblage.

*Specialist Studies – Zooarchaeological Analyses

• Faunal analysis includes the identification, analysis, and interpretation of faunal remains or animal bones recovered from the I-95 project.
  • Faunal analysis allows archaeologists to understand the animal life that existed in specific time periods on a given landscape and the culinary habits or foodways of the people who lived on that landscape.
  • Topics frequently covered by the analysis of faunal remains include:
    • Animal husbandry
    • Animal butchering techniques
    • Precontact and historic diet analysis and interpretation
    • Taphonomic process analysis

*Other Sample Types

• Archaeological samples have also been submitted for the following types of analysis:
  • AMS Radiocarbon Dating: Technique used to determine the age of organic materials recovered from archaeological sites.
  • Phytolith and Starch Grain Analysis: Analysis used to determine what types of plants were used by site inhabitants.
    • Artifacts and soil are submitted to be microscopically examined for plant silica bodies (phytoliths) or starch grains.
• Protein Residue Analysis: Analysis used to determine what types of animals were exploited by site inhabitants through examination of blood protein residues left behind on artifacts.
• AMS Radiocarbon samples are submitted to Beta Analytic.
• Phytolith/Starch Grain and Protein Residue Analysis samples are submitted to PaleoResearch.

Interpretation and Analysis

The principle unit of analysis and reporting for the I-95 Girard Avenue Interchange Project was that of the archaeological site. While not all of the work performed as part of the project falls under the umbrella of an archaeological site, the vast majority of the archaeological features and deposits requiring interpretation were located within an archaeological site boundary. While the archaeological site was the core interpretive unit for the project reporting and analysis, determining the extents of archaeological site boundaries in a complex urban landscape was an involved process that required repeated consultation with PennDOT and the PHMC.

Site Boundaries

For historical and many multicomponent sites, the city block within which individual lots/properties subjected to archaeological investigation were located was designated as the archaeological site, regardless of how many of the lots/properties within the block were investigated. The decision to incorporate the entire block within a historical site boundary was made in consultation with the PHMC. While PHMC guidelines specify that a site should encompass the full historical property boundary where possible, the frequently changing boundaries of historical properties and the urban setting raised questions about how to define a historical property. Furthermore, the PHMC recognized the unique concerns a project of this size and scale raises. Logistically, the urban setting and extensive sampling of the I-95 Girard Avenue Interchange Improvement Project would result in the creation of hundreds of individual sites that would be difficult to manage. Practically, the city block–level site designation would facilitate the addition of resources that might result from later construction monitoring. Interpretively, the PHMC concluded that deposits of similar age and survivability are likely to exist throughout a city block and that a site boundary should not therefore be determined by the limits of the APE or the limits of excavation imposed by that APE. Designating the entire city block as the site provided a measure of visibility for future planning.

As a general rule, the boundaries for pre-contact sites were typically determined by the physical extent of a site within the area investigated. This approach was applied in several cases across the I-95 Girard Avenue Interchange Improvement Project area; however, in many cases, pre-contact deposits and features survived in remnant Ap horizons that also contained historical materials and features. Given the overlapping nature of occupations in the urban environment, few sites were purely pre-contact. Based on consultation with the PHMC, it was determined that for locations where historical and pre-contact deposits occupied the same area of ground, the site would be classed as a multicomponent site, with its boundary being the extent that encompassed the totality of both resources. In practical terms, this meant that for most multicomponent sites, the historical boundary, being the largest area, served as the site boundary, with the location of the pre-contact deposits within that boundary being identified as loci within the PASS registration.

Reporting

The results of cultural resource investigations for the I-95 Girard Avenue Interchange Improvement Project are presented in multiple venues and formats. In addition to the products listed below, AECOM maintains an archive of project documentation, field notes and forms, monitoring reports, interim fieldwork summaries, and spatial data in their archaeological laboratory in Burlington, New Jersey. AECOM’s commitment to public outreach and involvement has resulted in numerous “pop-up” artifact displays to engage the project’s neighbors; temporary museum exhibits; a journal, River Chronicles; and the I-95 Girard Avenue Interchange Archaeology Center, a field laboratory and interpretive center with regular public hours.

Digging I-95

The project’s website, Digging I-95 (www.diggingi95.com), is the primary means by which information about the investigations is disseminated. This site is the clearinghouse for all public information about the archaeological and cultural work performed as part of the I-95 Girard Avenue Interchange Improvement Project, including interactive site reports, artifact galleries, neighborhood histories, and overarching neighborhood interpretations derived from archaeological excavation data. The site also includes a wealth of research about the material culture recovered during archaeological investigations, including an extensive visual archive of both 2D and 3D images of artifacts.

River Chronicles

This glossy, highly illustrated journal is intended for a general audience and presents summaries of archaeological and historical research of broad interest to the public. The journal provides a vehicle to communicate case studies and historical local color to the interested community, allowing for greater public engagement than what is typically part of a cultural resource management undertaking.

Site Reports

The most technical and detailed reporting was reserved for archaeological sites, which were designated in areas where investigations demonstrated the presence of intact deposits and/or cultural features. The classification of an excavated area as a site was typically predicated on the documentation of the survival of intact archaeological deposits and features possessing sufficient integrity to offer significant insight into past events and the lives of those who historically lived there. Areas with limited or no significant archaeological deposits were classified as non-sites and were not investigated further.

Prioritization of Parcels

The prioritization of parcels was intrinsically linked to the prioritization of the archaeological features encountered on a parcel. Parcels that occurred within a site or survey area but yielded no archaeological features or were not investigated were classed as low or no priority, respectively, and were not included in detailed analysis during reporting. Parcels that contained one or more high-priority features were considered high-priority parcels. Such parcels were examined in detail, including a full chain of title and property history research. This background research was done to help make connections between historical individuals who lived on a parcel and the archaeological materials recovered from features and deposits. The increased research into the property helped to enable more effective analysis and interpretation.

Non-Site Report

As the I-95 Girard Avenue Interchange Improvement Project area spanned four neighborhoods and hundreds of properties, resulting in the identification of dozens of archaeological sites, it was not possible to fully analyze each individual property. Investigations in non-site areas are documented in field notes and monitoring reports and will ultimately be discussed in a project-wide non-site report. Individual archaeological sites were documented in comprehensive site reports. In these reports, the archaeological findings of the entire site are discussed by property, with the most comprehensive discussions being reserved for the parcels containing high-priority archaeological features.

Neighborhood Syntheses

Data collected for the project will be analyzed at the neighborhood level and synthesized in a collection of neighborhood interpretive reports compiled and presented on the Digging I-95 website. Each neighborhood report (Port Richmond, Fishtown/Kensington, Northern Liberties, Delaware Wards, and the Landscape of the Lenape) will focus on a variety of thematic topics to enable comparison between sites within each neighborhood and help characterize these areas of the city in the past. Examples of neighborhood-level themes include population and demographics, foodways, health and sanitation, women, children, leisure, and alcohol consumption. The thematic approach will provide a framework for comparing similar features across sites within a neighborhood and will distill the plethora of archaeological data presented in the individual site reports into a manageable narrative of the neighborhood and its residents as seen through the lens of archaeology.